Changes in the cohort composition of turner syndrome and severe non-diagnosis of Klinefelter, 47,XXX and 47,XYY syndrome: a nationwide cohort study

Sex chromosome abnormalities (SCAs) - Turner syndrome (TS [45,X]), Klinefelter syndrome (KS [47,XXY]), Triple X syndrome (Triple X, [47,XXX]) and Double Y syndrome (Double Y,[47,XYY]) - are estimated to affect approximately 1 per 400 births [1]. A number of representative cytogenetic surveys were conducted years ago on newborns in various countries. Previously, we pooled data from these surveys in order to estimate the prevalence of TS, KS, Triple X and Double Y [2–5] which were; TS: 50 per 100,000 newborn females [24 TS among 48,744 newborn females] [6–10]; KS: 152 per 100,000 newborn males [84 KS among 55,212 newborn males] [6, 7, 10–13]; Triple X: 84 per 100,000 newborn females [62 Triple X among 73,990 newborn females] [4]; 4) Double Y: 98 per 100,000 newborn males [51 Double Y among 52,004 newborn males] [7, 10, 12, 14, 15]. Estimates are, however, subject to much uncertainty.

The presence of a SCA may affect individuals at multiple organ levels, but the range of affection is very wide. Abnormal and delayed puberty as well as infertility are, however, distinctive features in individuals affected by TS or KS [16, 17]. Previously, we and others, have reported an almost four-fold increased mortality [3–5, 18, 19] as well as increased morbidity associated to a wide range of diseases [18, 20] in individuals affected by SCAs compared to age and sex-matched controls. Further, we have reported a reduced socioeconomic status in individuals affected by a SCA with lower education (except for TS), increased risk of being retired from the labor market, lower income, and reduced likelihood of living in a relationship [21–24].

Delayed or even non-diagnosis of SCAs is common. Individuals not diagnosed in infancy often do not receive a diagnosis until years after relevant medical therapy should have been initiated to alleviate symptoms [4, 5, 25–27]. Further, delayed diagnosis prevents timely screening and intervention for common associated health problems as well as learning and behavioral disabilities [28, 29]. Thus, health related and social consequences of non-diagnosis and delayed diagnosis are a continuous concern.

In the present study we aimed to investigate change over time in incidence, prevalence, and age at diagnosis in a national cohorts of SCAs.

During 1961–2014 a total of 2875 individuals were diagnosed and recorded in the DCCR with a karyotype compatible with TS (n = 1156); KS (n = 1235); Triple X (n = 197); or Double Y (n = 287) (Table 1). Year of diagnosis did not differ among the syndromes (P = 0.07).

Population-based prevalence

The population-based prevalence of TS, KS, Triple X and Double Y increased during the study period (Fig. 1). Compared to the expected population-based prevalence, 70% of TS (980 out of 1418); 23% of KS (962 out of 4244); 7% of Triple X (165 out of 2381); and 9% of Double Y (239 out of 2736) were diagnosed and alive by the end of the study.

Linear extrapolation of the observed and expected population-based prevalence of TS show that these theoretically equals each other in 2039 – in other words, all live TS will be diagnosed in 2039, given the current diagnostic practices continue unaltered, given the expected population-based prevalence is correct, and given the mortality rate is identical in TS and in the background population. For KS, this point in time was estimated to the year of 2471. Theoretically, the expected and observed prevalence of Triple X and Double Y will continue to diverge owing to a greater increase in the expected prevalence than in the observed prevalence.

Incidence

During the study period an average of 2,648,000 females in Denmark were at risk of being diagnosed with a SCA. The average annual number of diagnosed TS was 24. Thus, the average incidence of TS was 9.0 per million females. The average number of 45,X TS and TS with other karyotypes than 45,X (“other” TS) diagnosed annually was 8 and 16, respectively, leading to an average incidence of 3.1 45,X TS and 5.9 “other” TS per million females. The incidence among all TS was stable during the study period (P = 0.10), whereas the incidence was decreasing for 45,X TS (P = 0.0006) and increasing for “other” TS (P < 0.0001) (Fig. 2).

An average of four Triple X was diagnosed each year during the study period leading to an incidence of 1.6 Triple X per million females. No change in incidence among Triple X was observed during the study period (P = 0.22) (Fig. 3a).

The average number of males in Denmark being at risk of being diagnosed with a SCA each year during 1970–2014 was 2,589,000, and annually an average of 24 KS and 6 Double Y were diagnosed. Thus, the average incidence of KS and Double Y were 9.4 and 2.2 per million males during the study period. The incidence was increasing for KS (P = 0.02) as well as for Double Y (P = 0.03) during the study period (Fig. 3b and c).

Prevalence among newborns

Median year of birth among the SCAs was as follows: 1) TS: 1971 (range: 1885–2014); 2) Triple X: 1975 (range: 1895–2014); 3) KS: 1965 (range: 1882–2013); and 4) Double Y: 1977 (range: 1899–2013). During 1901 to 2014, prevalence was divided into sub-periods according to differences in average prevalence. During 1961–1985 the maximum average prevalence of all TS was 59 per 100,000 newborn females (45,X TS: 21 per 100,000 newborn females and “other” TS: 38 per 100,000 newborn females), thus higher than expected. During 1971–1990 the maximum average prevalence of Triple X was 11 per 100,000 newborn females, corresponding to 13% of the expected prevalence. The maximum average prevalence of KS was observed during 1961–1990 being 57 per 100,000 newborn males, corresponding to 38% of the expected prevalence. Double Y had a maximum average prevalence of 18 per 100,000 newborn males during 1976–1990, corresponding to 18% of the expected prevalence (Fig. 4).

Age at diagnosis

The median age at diagnosis was for TS 15.1 (range: 0.0–85.4) years, for Triple X 17.9 (0.0–73.2) years, for KS 27.5 (0–82.8) years, and for Double Y 15.1 (0–70.7) years (Fig. 5a and b). 45,X TS was diagnosed with significantly less delay than TS with other karyotypes (median age: 45,X TS = 11.4 years versus “other” TS = 19.0 years) (P < 0.0001) (Fig. 5b). There was no change in age at diagnosis during 1970–2014 among all TS (P = 0.17), whereas age at diagnosis was decreasing among 45,X TS (P = 0.005). Age at diagnosis among TS with other karyotypes was stable (P = 0.99) during the study period, as well as among the other SCAs (Triple X: P = 0.28; KS: P = 0.72; and Double Y: P = 0.33). It is clear from the violin plots that the pattern of age at diagnosis is very different among different groups of SCAs (Fig. 5), with most KS being diagnosed much later than most TS, and with Triple X and Double Y being intermediate.

This population based study shows that the previously estimated prevalence of 50 TS per 100,000 may be an underestimate as we here present data showing a prevalence of 59 TS per 100,000 newborn females, corresponding to 1 per 1700. The karyotypic composition of the TS population is seemingly changing as the incidence of 45,X TS is decreasing and the incidence of TS with other karyotypes is increasing. Among KS and Double Y incidence is increasing as well, whereas it is stable among Triple X. Non-diagnosis remains extensive among KS, Triple X and Double Y. All TS eventually become diagnosed, although with a considerable delay.

The prevalence of the four forms of SCAs varied with a similar pattern during the study period. The marked increase in prevalence observed during the 1960’ties to the 1990’ties, possibly reflects that affected individuals from these birth cohorts were either born or reached puberty or fertil age at a time when karyotyping had become an established diagnostic procedure. However, comparing the maxiumum average prevalence with the expected prevalence of SCAs, the majority of KS (62%), Triple X (87%) and Double Y (82%) remain undiagnosed. In 2003, we reported that approximately 75% of expected KS was undiagnosed [2]. Although more with KS are diagnosed presently, we consider this far from satisfactory. In contrast, the average prevalence of TS exceeded the expected prevalence.

The population-based prevalence was steadily increasing for all SCAs during the study period caused by the build-up of the DCCR with continuing recruitment exceeding the rate of exit, a phenomenon known from other rare conditions as well [32]. The population-based prevalence will be stable when recruitment equals exit (death or emigration). Owing to increased mortality as well as increased rates of induced abortions among SCAs [33, 34] it will though remain lower than expected even with complete diagnosis of SCAs. However, it is difficult to ascertain how much prevalence is affected hereby. Only approximately 5% of pregnant women in Denmark have a prenatal karyotype performed [35], but they are of course selected as high-risk patients based on triple screening and non-invasive prenatal testing.

The incidence among all TS was stable during the study period. Interestingly, the incidence was steeply decreasing for 45,X TS and increasing for TS with other karyotypes. If this development continues it seems likely that 45,X TS are becoming extinct. Between 2004 and 2006, Denmark instituted a free prenatal screening program for Down syndrome (DS), in which over 90% of pregnant women in Denmark participate, and previously we have reported that approximately 40% of expected TS fetuses are detected by the DS screening [34]. The high induced abortion rate among prenatally detected 45,X TS fetuses likely contribute to the decreasing incidence of 45,X TS. Induced abortion among TS fetuses with other karyotypes are less pronounced [34], and a generally more favorable phenotype among these TS [36, 37] may prevent early diagnosis. A likely explanation for the increase in incidence among TS with other karyotypes may be that fertility treatment has become more common and thus more are diagnosed owing to fertility problems.

Late diagnosis as well as non-diagnosis of SCAs have been a continuous concern [25–27, 38] since SCAs are associated with increased morbidity and mortality [3, 5, 20, 23, 39], learning and/or behavioral disabilities [40–45] as well as reduced socioeconomic outcomes. Although there is lack of evidence regarding the influence of age at diagnosis on long-term outcomes, we believe that early diagnosis will provide better overall long-term outcome in SCAs by providing an opportunity for timely intervention against associated health problems as well as against learning and behavioral problems. Regrettably, the present study shows that delay in diagnosis remains a major problem (Fig. 5). Increased vigilance among health care professionals as well as a more liberal access to diagnostic tools was however hypothesized as having contributed to a decreasing diagnostic delay. As previously suggested for both TS and KS, neonatal screening programs would allow complete diagnosis of SCAs without diagnostic delay. Population-based, neonatal screening can be considered for conditions which have: 1) the magnitude to be an important health problem with a latent and early asymptomatic stage; and 2) has a well-understood natural history for which there are accepted treatments with associated facilities for diagnoses and treatment [46]. We consider these requirements fulfilled for SCAs. Due to the rarity of the syndromes it will, however, take a long time to demonstrate associations between early diagnosis, continuous specialized care and improved long-term outcomes.

The broad clinical spectrum observed among SCAs, ranging from overt to minimal or no apparent clinical features and only subtle symptoms [17, 47, 48] likely relates to the non-diagnosis and the delayed diagnosis. The fate of individuals with SCAs escaping diagnosis remains an enigma. Possibly, undiagnosed individuals experience a spectrum of similar challenges as those diagnosed, yet they might remain undiagnosed due to reluctance of seeking medical advice or due to lack of attention from health professionals. Based on our clinical experience, we believe that individuals diagnosed late in life more or less struggle with similar problems as those being diagnosed early in life. Recent data from UK among TS and Triple X support this hypothesis [36]. Here a large number of presumably undiagnosed 45,X and 47,XXX females were detected approximately about 250,000 examined females, presenting with typical features for both of these syndromes.

The karyotypic composition of TS is changing with less 45,X and more mosaic TS being diagnosed, possibly due to increased frequency of legal abortions among 45,X TS. TS have a higher prevalence than previously anticipated as observed in 1 per 1700 newborn females. The majority of KS, Triple X and Double Y remain undiagnosed despite an increase in diagnostic activity among KS and Double Y. Delayed diagnosis is a continuous problem among all SCAs, and no change over time has been observed. We consider a newborn screening program as the only opportunity to reduce both diagnostic delay as well as non-diagnosis and with a potential to improve health and socioeconomics among individuals affected by SCAs.