3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 853
ORIGINAL RESEARCH
The Thy-Box for Sonographic
Assessment of the Fetal Thymus
Nomogram and Review of the Literature
Mar Muñoz-Chápuli, MD, Francisco Gámez, MD, PhD, Coral Bravo, MD, PhD, Luis Ortiz, MD, PhD,
Ricardo Pérez, MD, PhD, Juan A. De León-Luis, MD, PhD
Received June 19, 2014, from the Fetal Medicine
Unit, Department of Obstetrics and Gynecology,
Hospital General Universitario Gregorio Marañón,
Universidad Complutense de Madrid, Madrid,
Spain (M.M-C., F.G., C.B., L.O., R.P., J.A.D.L.-
L.); and Department of Obstetrics and Gynecol-
ogy, Hospital Central de la Defensa Gómez Ulla,
Universidad de Alcalá de Henares, Madrid, Spain
(C.B.). Revision requested July 28, 2014. Revised
manuscript accepted for publication August 25,
2014.
We thank Almudena Valdés for assistance
with manuscript preparation. This work was sup-
ported by the Fondo de Investigaciones Sanitarias
(grant PI13/02769).
Address correspondence to Juan A. De León-
Luis, MD, PhD, Fetal Medicine Unit, Department
of Obstetrics and Gynecology, Hospital General
Universitario Gregorio Marañón, Calle O’Don-
nell 48, Planta 0, Bloque C, 28009 Madrid, Spain.
E-mail: jdeleonluis@yahoo.es
Abbreviations
GA, gestational age; ICC, intraclass correla-
tion coefficient
doi:10.7863/ultra.34.5.853
©2015 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2015; 34:853–858 | 0278-4297 | www.aium.org
Objectives—To assess the feasibility and reproducibility of fetal thy-box visualization
and to set normative data for the fetal thymus using this technique.
Methods—We performed a cross-sectional observational study in pregnant women in
their second trimester of pregnancy who attended the fetal medicine unit of the Hos-
pital Gregorio Marañón from March 2011 to March 2013. Using thy-box sonography,
which sets boundaries within the thymus, we assessed the feasibility of the thy-box and
analyzed the factors related to its visualization. Measurements of the thy-box antero-
posterior and transverse diameters were performed in healthy fetuses. Interobserver
agreement was studied for both items. We set normative data for the thy-box diameters
in singleton and twin pregnancies.
Results—A total of 337 patients from 15 to 37 weeks were recruited after exclusion cri-
teria. Thy-box feasibility was 74.2% (250 of 337). Nonfeasible cases were related to the
fetal anterior spine and advanced gestational age. Assessment of the thy-box antero-
posterior and transverse diameters was achievable in 250 fetuses. Interobserver agree-
ment was good for feasibility and measurements (κ = 0.80). Linear growth of the
thy-box in both diameters was observed in relation to gestational age during the sec-
ond half of pregnancy. Thy-box diameters did not show significant differences accord-
ing to the type of gestation, fetal sex, or chorionicity.
Conclusions—Thy-box diameters are achievable and reproducible by sonography in
normal singleton and twin pregnancies during the second trimester; however, although
feasible, the thy-box transverse diameter measurement is not reliable beyond 28 weeks.
Further studies are needed to evaluate thy-box diameters in pathologic cases, especially
conotruncal anomalies.
Key Words—fetal thymus; nomogram; obstetric ultrasound; prenatal sonography; thy-
box; thymus measurements
T
he thymus gland, which is located in the cranial and ventral
portions of the mediastinum, has a determinant role in the
development of competent cellular immunity.1 The finding
of thymus agenesis or hypoplasia has been associated with 22q11
microdeletion in the presence of conotruncal anomalies.2 Thymic
hypoplasia is also a common finding in cases of preterm prelabor
rupture of membranes, intrauterine growth restriction, maternal
preeclampsia, and fetuses affected with Down syndrome and other
aneuploidies, which may explain the immune function impairment
in those cases.3–10
3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 854
Muñoz-Chápuli et al—The Thy-Box for Sonographic Assessment of the Fetal Thymus
Sonographic assessment of the fetal thymus can be
performed by different methods2,11–15; however, it still
presents some difficulties.6 The perimeter is not always
easily defined and occasionally requires too much time.11
The thymic transverse diameter has the extra difficulty of
the subtle echogenic interphase between the thymus
and the lungs, which also varies with gestational age
(GA).12 The anteroposterior diameter may be narrowed
in cases of conotruncal anomalies,13 as the 3 vessels can be
located ventrally. These difficulties set the need for a com-
mon anatomic landmark: the thy-box.14,16
The thy-box, described by Paladini,14 sets the bound-
aries within the thymus and can be easily assessed.
Limited on both lateral sides by the internal mammary
arteries, on the posterior side by the ductal and aortic arch,
on the ventral side by the sternum, and depicted with color
Doppler imaging, the thy-box may be an option for repro-
ducible evaluation of the thymus (Figure 1). The aim of our
study was to evaluate the feasibility and reproducibility of
the thymus visualization in healthy singleton and twin
fetuses from 15 to 37 weeks’ gestation using the thy-box
sonographic view and to set normative data for thymus
diameters with this technique.
Materials and Methods
We conducted a cross-sectional observational study in the
Fetal Medicine Unit of the Hospital Gregorio Marañón
between March 2011 and March 2013. The local Ethics
Committee approved the research protocol, and informed
consent was obtained from all patients. Patients were
recruited consecutively from a cohort of pregnant women
referred for a routine prenatal sonographic evaluation.
All women had their pregnancies dated during first-
trimester sonography; in the case of twins, chorionicity was
also determined in this scan. Inclusion criteria were a sono-
graphically confirmed last menstrual period and absence
of major fetal anomalies. Cases affected by intrauterine
growth restriction and twin-twin transfusion syndrome were
excluded prenatally. Furthermore, cases were excluded post-
natally because of preterm birth, adverse perinatal out-
comes, and loss of follow-up.
Sonographic examinations were performed by an
expert sonographer (F.G.) using 2 ultrasound systems with
transabdominal 4–8-MHz transducers (Voluson Expert;
GE Healthcare, Zipf, Austria; and Aplio 300; Toshiba
Medical Systems Co, Ltd, Tokyo, Japan). In our sono-
graphic examinations, fetal biometry was assessed by the
biparietal diameter, abdominal circumference, and femur
length. We evaluated the thymus in the thy-box axial view
854
described by Paladini14 and measured the fetal thymic
anteroposterior and transverse diameters in the thy-box
view (Figure 1).
To assess interobserver agreement, 39 extra cases
covering all of the gestational weeks in the study were
randomly reevaluated by another sonographer (C.B.).
The thymus was measured independently by 2 observers;
neither of them knew about the results obtained by the
other. The agreement was interpreted according to
the criteria of Landis and Koch17 as poor (intraclass
correlation coefficient [ICC], <0.21), fair (ICC, 0.21–
0.40), moderate (ICC, 0.41–0.60), good (ICC, 0.61–0.80),
or excellent (ICC, >0.80).
The feasibility of thy-box visualization and measure-
ments and their relationship with maternal and fetal vari-
ables (body mass index, uterine anomalies such as uterine
fibroids and congenital malformations, type of pregnancy,
and fetal position) and GA were assessed. P < .05 was con-
sidered statistically significant.
For elaborating the nomograms, we followed the
method described by Royston and Wright.18 First, we
selected the lowest polynomial degree (explored up to
the third degree) that was best adapted to the data using
the least squares method. Then we explored the residual
data for any changes in the assumptions of normality and
homoscedasticity. The latter was obtained by using previ-
ous estimates for scaled absolute residuals and a regression
model of these residuals over weeks’ gestation. For inho-
mogeneous variance residuals, we chose to model the scaled
absolute residuals so they could be incorporated into the
intervals of normality. Finally, we estimated the 95% con-
fidence intervals for each week of pregnancy.
The relationship between thymic measurements and
GA was also studied according to the type of pregnancy
(singleton versus bichorionic/monochorionic twins) and
fetal sex, which was assessed by sonography and checked
postnatally. By a linear regression model, we extracted the
predicted means and standard deviations of the thymic
measurements for each gestational week. Data were ana-
lyzed with SPSS version 21 software (IBM Corporation,
Armonk, NY).
Results
A total of 529 fetuses were initially included in our study;
192 cases were excluded due to loss of follow-up (19.0%),
preterm birth (13.4%), and adverse perinatal outcomes
(3.0%). A total of 337 fetuses formed the final cohort after
the exclusion criteria. Clinical and sonographic character-
istics of our cohort are shown in Table 1.
J Ultrasound Med 2015; 34:853–858
3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 855

Muñoz-Chápuli et al—The Thy-Box for Sonographic Assessment of the Fetal Thymus

Visualization and measurement of both the antero-
posterior and transverse diameters in the thy-box were
possible in 250 cases (74.2%). In all cases, both diameters
were equally observed. The interobserver agreement results
for the feasibility of the thy-box showed a κ value of 0.80.
Interobserver agreement for the thy-box measurements
showed high concordance between observers. The ICC
values were 0.81 (95% confidence interval, 0.556–0.923)
for the anteroposterior diameter and 0.87 (95% confidence
interval, 0.715–0.960) for the transverse diameter.
The main factors related to an impossibility of assess-
ing the measurements were an adverse fetal position (anterior
spine) and advanced GA. Nonfeasibility of the measurements
varied from 60.8% in cases of a fetal anterior spine to 19.8%
in any other fetal position (relative risk, 3.07; P < .001).
Gestational age showed an inverse relationship with the
possibility of achieving thymus measurements (P = .006).
We also found that after 28 weeks’ gestation, the accuracy
Figure 1. View of the thy-box in a 23-week-old fetus. Dashed blue lines
show thy-box diameters (A, transverse; B, anteroposterior). IMA
indicates internal mammary artery; IV, innominate vein; T, trachea; and
Thy, thymus.
of the thy-box anatomic frame was limited, as the interphase
between the thymic parenchyma and the lungs seemed to
exceed the lateral limits of the thy-box, which were the
internal mammary arteries (Figure 2). This finding led us
to truncate the values of the thy-box diameters between 15
and 26 weeks’ gestation.
No other maternal or fetal factors were associated with
lower feasibility. We found no significant differences in the
feasibility of thy-box measurements regarding the type of
pregnancy (singleton versus twins; P = .48), the presence
of uterine fibroids or uterine malformations (P = .44), or
the examination time (P = .08).
We also assessed thymic anteroposterior and trans-
verse diameters and studied their development through-
out gestation. Figure 3 shows the means and standard
deviations of the anteroposterior and transverse diameters.
The predicted mean of the anteroposterior diameter was
calculated as anteroposterior diameter = 0.32 × GA – 1.23
(R2 = 0.26; P < .001; SD = 1.35). The predicted mean of
the transverse diameter was calculated as transverse diam-
eter = 0.84 × GA – 6.48 (R2 = 0.73; P < .001; SD = 1.33).
Table 2 shows the predicted means and 5th and 95th
percentiles for normal thy-box diameters at each GA.
These results did not differ significantly depending on
the type of pregnancy (singleton versus twins; Figure 4),
chorionicity in twin pregnancies (P = .33), or the fetal sex
(P = .54).

Discussion

In this study, we found that the thy-box can be visualized

in both singleton and twin pregnancies, just as with other
methods of sonographic evaluation of the fetal thymus.11,12,15
The feasibility of this technique was high, as it was possible
Table 1. Clinical and Sonographic Characteristics of the Study to achieve visualization in three-fourths of the cases. Its repro-
Population, Including Perinatal Outcome Data ducibility between observers was also high, which con-
Characteristic Value
tributed to the reliability of the method.
However, we found that, although feasibility was high
Age, y 34 ± 4.68 and similar to that of other methods,16 this method still
Body mass index, kg/m2 24.5 ± 3.38
Uterine anomalies, n (%) 14 (4.1)
presents some difficulties, especially in cases of adverse fetal
Singleton/twin pregnancy, n (%) 246 (73.0)/90 (27.0) positions during the scan. When the fetal spine is located
Monochorionic/dichorionic, n (%) 22 (24.4)/68 (25.6) right below the transducer, it interposes between the ultra-
Anterior fetal spine/other positions, n (%) 62 (18.4)/275 (81.6) sonic beam and the aim of the scan, producing a shadow that
Examination time, min 5.8 ± 3.9 hides the limits of this organ. The fetal position becomes
GA at examination, wk 19.9 ± 4.1
Biparietal diameter, mm 46.7 ± 8.6
especially important because of the need for color Doppler
Abdominal circumference, mm 151.7 ± 38.8 imaging for proper observation of the lateral limits of the
Femur length, mm 31.4 ± 8.1 thy-box: the internal mammary arteries. An inadequate
GA at birth, wk 38.8 ± 1.61 fetal position prevents the vessels from being filled by color
Birth weight, g 3084 ± 503 and thus being observed. Other maternal factors, such as
Data are presented as mean ± SD where applicable. the body mass index and uterine anomalies, were not

J Ultrasound Med 2015; 34:853–858 855

3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 856

Muñoz-Chápuli et al—The Thy-Box for Sonographic Assessment of the Fetal Thymus

related to a decrease in feasibility, similar to other studies
that assessed adequate visualization of cardiac anatomy in
transverse views in the second trimester.19,20 On the other
hand, advancing GA seems to cause a decrease in the per-
centage of thy-box visualization, probably secondary to a
lesser delineation of its limits, as well as impairment of the
real contours of the gland and the lateral limits of the thy-
box (Figure 2).
In this study, we have shown that thy-box diameters
can be obtained by sonography in the second and third
trimesters of pregnancy. Both thy-box measurements
showed excellent reproducibility in the interobserver
study, according to the criteria of Landis and Koch.17
As described in previous studies, this gland develops in
a linear pattern during fetal life,2,11–16 which was also con-
firmed in our study (Figure 3). No differences were found
between male and female fetuses or between singleton and
twin bichorionic and monochorionic pregnancies, and
these findings were consistent with previous data.16,21,22
The thymic transverse diameter also shows a higher variation
throughout pregnancy than the anteroposterior diameter,
which has been described with other techniques as well and
may be related to the anatomy of the mediastinum.2,12,16,21,23
According to the results of this study, the thy-box
seems to be a feasible and reproducible method not only
for qualitatively confirming the presence of the fetal thy-
mus but also for establishing its dimensions throughout
pregnancy. Nevertheless, this technique had certain limi-
tations. One of the most relevant seems to be that the thy-
box transverse diameter is not reliable in the third trimester,
as we noticed beyond 28 weeks’ gestation. In these cases,
the transverse diameter of the thymus exceeds the lateral
limits of the thy-box, which could substantially underesti-
mate the true transverse diameter of the gland (Figure 3).

Figure 2. Limitations of the thy-box transverse diameter (TD) in the third trimester. Left, Anatomic transverse thymus diameters in 28- and 30-week-
old fetuses. Right, Thy-box diameters (1, transverse; 2, anteroposterior). The true limits of the fetal thymus are underestimated by the thy-box trans-
verse diameters in the third trimester.

856 J Ultrasound Med 2015; 34:853–858

3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 857

Muñoz-Chápuli et al—The Thy-Box for Sonographic Assessment of the Fetal Thymus

Another factor that can modify the usefulness of the thy-
box is the need for color Doppler imaging, which means
that its assessment requires equipment that contains this
tool and sonographer knowledge. Although more and more
units are incorporating color Doppler imaging in routine
sonographic scans, especially for evaluation of the fetal
heart at 15 to 22 weeks, this method may still be reserved for
tertiary centers where cases at risk of thymic hypoplasia
(mainly conotruncal congenital heart disease) are evalu-
ated and managed.
The thy-box anatomic landmark may establish a con-
sensus among sonographers and standardize future studies.
Thymus evaluation during pregnancy is not performed
during routine sonography, but interest in it has increased.

Figure 3. Assessment of thy-box anteroposterior (A) and transverse Table 2. Predicted Means and Percentiles for Anteroposterior and
(B) diameters. Lines represent means and standard deviations. Transverse Diameters
Predicted mean formulas are also shown. APD indicates anteroposte-
GA, wk Mean 5th 95th
rior diameter; and TD, transverse diameter.
Anteroposterior
15 3.52 0.82 6.23
16 3.84 1.13 6.55
17 4.16 1.45 6.87
18 4.48 1.77 7.18
19 4.79 2.08 7.50
20 5.11 2.40 7.82
21 5.43 2.72 8.14
22 5.74 3.03 8.45
23 6.06 3.35 8.77
24 6.38 3.67 9.09
25 6.69 3.99 9.40
26 7.01 4.30 9.72
27 7.33 4.62 10.04
Transverse
15 6.18 3.52 8.84
16 7.02 4.36 9.68
17 7.87 5.21 10.53
18 8.71 6.05 11.37
19 9.56 6.90 12.22
20 10.40 7.74 13.06
21 11.24 8.58 13.90
22 12.09 9.43 14.75
23 12.93 10.27 15.59
24 13.78 11.12 16.44
25 14.62 11.96 17.28
26 15.46 12.80 18.12
27 16.31 13.65 18.97

Figure 4. Relationship between thy-box transverse diameter and GA in

singleton and twin pregnancies.

J Ultrasound Med 2015; 34:853–858 857

3405jum749-894online_Layout 1 4/22/15 8:11 AM Page 858
Muñoz-Chápuli et al—The Thy-Box for Sonographic Assessment of the Fetal Thymus
As Chaoui et al2 established, among fetuses with conotrun-
cal anomalies, thymic hypoplasia or aplasia has sensitivity
of 90.0%, specificity of 98.5%, a positive predictive value of
81.8%, and a negative predictive value of 99.2% for identi-
fication of fetuses with 22q11 microdeletion. Moreover,
there is an association between fetal thymic hypoplasia and
obstetric disorders such as fetal inflammatory response
syndrome in the context of prolonged premature rupture
of membranes, intrauterine growth restriction, maternal
preeclampsia, and chromosomal abnormalities such as
Down syndrome, suggesting the role of the immune system
in these disoders.3–10 Under these conditions, sonographic
assessment of the fetal thymus by the thy-box or any other
method becomes extremely important. Future studies are
needed to assess thy-box diameters in cases of pathologic
fetal findings, especially conotruncal anomalies and other
obstetric and perinatal disorders associated with thymus
hypoplasia.2,16
In conclusion, we have shown that visualization of the
thy-box and determination of the anteroposterior and
transverse diameters is feasible between 15 and 27 weeks
of pregnancy, with excellent interobserver reproducibility.
The major limitation to its use seems to be an adverse fetal
position. The effectiveness of the thy-box is limited in the
third trimester, as the transverse diameter of the gland is
underestimated. Thy-box measurements could be useful
for sonographic evaluations of fetal thymic development
and for studies of this gland in pathologic conditions dur-
ing the second trimester. Further studies are needed to
evaluate thy-box diameters in pathologic cases.
References
1. Di George AM, Lischner HW, Dacou C, Arey JB. Absence of the thymus.
Lancet 1967; 1:1387.
2. Chaoui R, Heling KS, Sarut Lopez A, Thiel G, Karl K. The thymic-thoracic
ratio in fetal heart defects: a simple way to identify fetuses at high risk for
microdeletion 22q11. Ultrasound Obstet Gynecol 2011; 37:397–403.
3. Musilova I, Hornychova H, Kostal M, Jacobsson B, Kacerovsky M.
Ultrasound measurement of the transverse diameter of the fetal thymus
in pregnancies complicated by the preterm prelabor rupture of
membranes. J Clin Ultrasound 2013; 41: 283–289.
4. Yinon Y, Zalel Y, Weisz B, et al. Fetal thymus size as a predictor of
chorioamnionitis in women with preterm premature rupture of mem-
branes. Ultrasound Obstet Gynecol 2007; 29:639–643.
5. El-Haieg DO, Zidan AA, El-Nemr MM. The relationship between sono-
graphic fetal thymus size and the components of the systemic fetal
inflammatory response syndrome in women with preterm prelabour rup-
ture of membranes. BJOG 2008; 115:836–841.
6. Chromi A, Ghezzi F, Raffaelli R, Bergamini V, Siesto G, Bolis P.
Ultrasonographic measurement of thymus size in IUGR fetuses: a marker
858
of the fetal immunoendocrine response to malnutrition. Ultrasound Obstet
Gynecol 2009; 33:421–426.
7. Eviston DP, Quinton AE, Benzie RJ, Peek MJ, Martin A, Nanan RK.
Impaired fetal thymic growth precedes clinical preeclampsia: a case-
control study. J Reprod Immunol 2012; 94:183–189.
8. Mohamed M, Eviston DP, Quinton AE, et al. Smaller fetal thymuses in
pre-eclampsia: a prospective cross-sectional study. Ultrasound Obstet
Gynecol 2011; 37:410–415.
9. De León-Luis J, Santolaya J, Gámez F, Pintado P, Pérez R, Ortiz-Quintana
L. Sonographic thymic measurements in Down syndrome fetuses.
Prenat Diagn 2011; 31:841–845.
10. Karl K, Heling KS, Sarut Lopez A, Thiel G, Chaoui R. Thymic-thoracic
ratio in fetuses with trisomy 21, 18 or 13. Ultrasound Obstet Gynecol 2012;
40:412–417.
11. Zalel Y, Ganzu R, Mashiach S, Achiron R. The development of the fetal
thymus: an in utero sonographic evaluation. Prenat Diagn 2002; 22:114–
117.
12. Cho JY, Min JY, Lee IH, McCrindle B, Hornberger LK, Yoo SJ. Diameter
of the normal fetal thymus on ultrasound. Ultrasound Obstet Gynecol 2007;
29:634–638.
13. Felker RE, Cartier MS, Emerson DS, Brown DL. Ultrasound of the fetal
thymus. J Ultrasound Med 1989; 8:669–673.
14. Paladini D. How to identify the thymus in the fetus: the thy-box.
Ultrasound Obstet Gynecol 2011; 37:488–492.
15. Li L, Bahtiyar MO, Buhimschi CS, Zou L, Zhou QC, Copel JA.
Assessment of the fetal thymus by two- and three-dimensional ultrasound
during normal human gestation and in fetuses with congenital heart
defects. Ultrasound Obstet Gynecol 2011; 37:464–469.
16. Mu~noz-Chápuli M, Bravo Arribas C, Gámez-Alderete F, Fernández-
Pacheco RP, Ortiz-Quintana L, De León-Luis J. Ultrasound scan of the
fetal thymus [in Spanish]. Ginecol Obstet Mex 2014; 82:43–49.
17. Landis JR, Koch GG. The measurement of observer agreement for cate-
gorical data. Biometrics 1977; 33:159–174.
18. Royston P, Wright EM. How to construct “normal ranges” for fetal vari-
ables. Ultrasound Obstet Gynecol 1998; 11:30–38.
19. Schwarzler P, Senat MV, Holden D, Bernard JP, Masroor T, Ville Y.
Feasibility of the second-trimester fetal ultrasound examination in an uns-
elected population at 18, 20 or 22 weeks of pregnancy: a randomized trial.
Ultrasound Obstet Gynecol 1999; 14:92–97.
20. DeVore GR, Medearis AL, Bear MB, Horenstein J, Platt LD. Fetal
echocardiography: factors that influence imaging of the fetal heart during
the second trimester of pregnancy. J Ultrasound Med 1993; 12:659–663.
21. De León-Luis J, Gámez F, Pintado P, et al. Sonographic measurements of
the thymus in male and female fetuses. J Ultrasound Med 2009; 28:43–48.
22. Gámez F, De León-Luis J, Pintado P, et al. Fetal thymus size in uncom-
plicated twin and singleton pregnancies. Ultrasound Obstet Gynecol 2010;
36:302–307.
23. Gardiner H, Chaoui R, The fetal three-vessel and tracheal view revisited.
Semin Fetal Neonatal Med 2013; 18:261–268.
J Ultrasound Med 2015; 34:853–858