Prenatal Assessment of the Antero-Posterior Jaw Relationship in HumanFetuses: From Anatomical to Ultrasound Cephalometric Analysis

Guillaume Captier, M.D., Ph.D., Jean-Michel Faure, M.D., Marcel Baumler, M.D., Francois Canovas, M.D., Ph.D.,Christophe Demattei, Ph.D., Jean-Pierre Daures, M.D., Ph.D.

Objectives: We wished to develop an ultrasound cephalometric analysis,particularly of the antero-posterior jaw relationship, to increase the accuracy ofprenatal diagnosis of retrognathism during the routine midterm test.

Methods: Anatomical cephalometric analysis was performed in 18 formalin-fixed human fetuses (between 16 and 39 gestational weeks), and ultrasoundcephalometry was prospectively carried out in 52 pregnant women (21 to 25gestational weeks). The same landmarks were used in the anatomical andultrasound median sagittal planes for comparison. Four cephalometric angleswere measured relative to the anterior cranial base: alveolar projection of themaxilla and the mandible, chin projection, and facial angle. The antero-posteriorjaw discrepancy was calculated.

Results: The projection of the maxilla was similar in the two cephalometricanalyses (IC [23.39, 0.23]), whereas the values of the projection of the mandiblewere lower in the ultrasound sample. The slope of the regression line of theantero-posterior jaw discrepancy on fetuses’ age did not show significantdifferences (IC [20.05, 1.54]) between anatomical and ultrasound cephalometry,although a difference of 3.23u ± 0.78u (IC [1.69, 4.77]) was observed. Despite thisvariability, the projections of mandible and chin were well determined by theprojection of the maxilla both in the anatomical and ultrasound sample.

Conclusions: Cephalometric analysis by prenatal sonography can beperformed to study the antero-posterior jaw relationship. We think that thisprocedure could be useful to improve prenatal diagnosis of retrognathism inhigh-risk pregnancies. Further studies should address the reproducibility andaccuracy of such analysis.

KEY WORDS: cephalometry, cleft palate, cranial base, retrognathism, Robinsequence

Congenital retrognathism with or without Robin se-

quence is a heterogeneous condition that is due to

mandibular retrognathia (position) or mandibular micro-

gnathia (size) and can be isolated, syndromic, or associated

with other abnormalities. Chromosomal abnormalities are

detected in 66% of fetuses with micrognathia (Nicolaides et

al., 1993), and 93% of the cases of isolated micrognathia,

according to prenatal ultrasound diagnosis, have at least

one additional abnormality after neonatal examination

(Vettraino et al., 2003). The most frequently associated

abnormalities are respiratory difficulties at delivery, pres-

ence of cleft palate, and/or developmental delay often due

to feeding problems (Vettraino et al., 2003).

In many countries, prenatal screening for anomalies of

the amniotic fluid is included in the routine bidimensional

(2D) ultrasound scan carried out at mid gestation as it is

considered an easy and cost-effective test (Dommergues et

al., 2006). During this mid-gestation ultrasound scan, the

position of the chin is imaged in the fetal profile. If

retrognathism is suspected, several authors have advocated

Dr. Captier is Pediatric Plastic Surgeon and Anatomist, Universite

Montpellier 1, UFR Medecine, Laboratoire d’Anatomie, Montpellier,

France, and CHRU Montpellier, Unite de Chirurgie Plastique Pediatri-

que, Hopital Lapeyronie, Montpellier, France. Dr. Faure and Dr.

Baumler are Gynecologist Obstetricians, CHRU Montpellier, Departe-

ment de Gynecologie Obstetrique, Hopital Arnaud de Villeneuve,

Montpellier, France. Dr. Canovas is Orthopedic Surgeon and Anatomist,

Chief Departement, Universite Montpellier 1, UFR Medecine, Labor-

atoire d’Anatomie, Montpellier, France. Dr. Demattei, is Biostatistician

and Research Engineer, Department of Biostatistics and Epidemiology

(BESPIM), University Hospital, Nımes, France. Dr. Daures is Biostat-

istician and Chief Departement, Universite Montpellier 1, Epidemiologie

Biostatistiques et Sante Public, IURC, Montpellier, France.

This work has received financial support from the French Agence

Nationale de la Recherche, grant ANR-08-BLAN-0272-03. This work was

presented orally at Craniofacial Surgery: Controversies and Consensus,

ISCFS XII Biennial International Congress, September 26–30, 2009,

Oxford, United Kingdom.

Submitted November 2009; Accepted July 2010.

Address correspondence to: Dr. Guillaume Captier, Unite de Chirurgie

Plastique Pediatrique, Hopital Lapeyronie, 371 avenue du doyen Gaston

Giraud, 34000 Montpellier, France. E-mail g-captier@chu-montpellier.fr.

DOI: 10.1597/09-221

465

the use of 2D or 3D sonographic imaging for an accurate

biometric evaluation of the mandible and of the craniofa-

cial skeleton (Otto and Platt, 1991; Chitty et al., 1993;

Paladini et al., 1999; Rotten et al., 2002; Tsai et al., 2004;

Roelfsema et al., 2006; Zalel et al., 2006) as the distinction

between micrognathia and retrognathia is essential for the

prognosis. Nevertheless, the ultrasound prenatal diagnosis

of retrognathism remains difficult, and it is often subjective

and underestimates its frequency (Nicolaides et al., 1993).

Since Broadbent (1931), cephalometric x-ray is an essential

tool in orthodontics and maxillofacial surgery for evaluating

craniofacial growth. For the diagnosis of retrognathism, the

evaluation of the antero-posterior jaw relationship in the

sagittal plan is considered an indispensable step. This

relationship is generally determined by cephalometric anal-

ysis using several landmarks, such as the ANB angle, which is

one of the most used parameters (Chang, 1987; Oktay, 1991).

This is possible because the cephalograms (i.e., lateral

cephalometric head radiographs) used for the analysis are

reproducible thanks to the standardization of the technique.

Since in utero x-ray cephalometry cannot be performed,

we wanted to evaluate the feasibility of determining the

antero-posterior jaw relationship using images of the fetus

profile in the medial sagittal plane taken during the routine

mid-gestation ultrasound scan. To this aim, we have

compared cephalometric analyses performed using the

same landmarks and reference lines in anatomical cephalom-

etry of formalin-fixed fetuses and ultrasound images in the

median sagittal plane.

MATERIAL AND METHODS

Anatomical Sample

This study was conducted using 18 formalin-fixed human

fetuses taken randomly from the anatomy laboratory

collection. The preparation and preservation of these

fetuses were previously described (Captier et al., 2008).

There were 12 second-trimester fetuses at 16 to 24

gestational weeks (GW) and six third-trimester fetuses

aged between 30 and 39 GW. Measurements were acquired

directly from the digitized images of the anatomical median

sagittal plane (AMSP) of each side the fetal head using the

Corel Draw 8 vector analysis software (Corel Corp.,

Fremont, CA). All measurements were done by one

investigator (G.C.), and the mean value of the right and

left measurements was used for statistical analysis.

Ultrasound Sample

A transversal prospective study was conducted in normal

low-risk pregnancies. A group of 52 randomly chosen

pregnant women (mean, 29.4 years; range, 18 to 39 years)

who filled the criteria of singleton pregnancy, morpholog-

ically normal fetus, gestational age based on the sono-

graphic measurement of the crown-rump length in early

pregnancy, and estimated fetal biometrics within the mean

(Table 1) were included in accordance with the moral,

ethical, regulatory, and scientific principles governing

clinical research as set out in the Declaration of Helsinki.

There were 23 male, 17 female, and 12 fetuses of unknown

sex. Acquisition of the ultrasound median sagittal plane

was performed during the routine second-trimester sono-

graphic examination (i.e., usually between 21 and 25 GW),

and this took no extra time (Babcook et al., 1996).

All of the sonographic examinations were performed by

one investigator (J.-M.F.) who is a national expert in

prenatal diagnosis and specialized in the exploration of the

fetal head (Faure, Baumler, et al., 2007; Faure, Captier, et

al., 2007; Faure et al., 2008). In each case, an ultrasound

image in the median sagittal plane (UMSP) was acquired

and measurements were directly performed using a General

Electric Voluson 730 Expert BT03 apparatus (GE Medical

Systems Kretz, Zipf, Austria).

Landmarks and Measurements

The antero-posterior jaw relationships in the AMSP and

in the UMSP were determined using five main landmarks:

the pituitary point, the nasion, the prosthion, the infra-

dental, and the gnathion.

The pituitary point (P) is the midline point on the raised

tuberculum sellae of the body of the sphenoid (Fig. 1). It is

located between the basipresphenoid and basipostsphenoid

at the level of the midsphenoid synchondrosis. In the fetus,

this synchondrosis is not ossified and can thus be identified

in the UMSP as a hypoechogenic dark line between the

ossification center of the basipresphenoid and the basipost-

sphenoid. P is at the endocranial extremity of this dark line

(Fig. 2). The nasion (N) is a facial landmark located at the

naso-frontal suture. The prosthion (A9) is the inferior

extremity of the maxillary alveolar process that corre-

sponds to the apex of the upper gum in the fetus. The

infradental (B9) is the superior extremity of the mandibular

alveolar process, which corresponds to the apex of the

TABLE 1 Biometric Data of the Ultrasound Sample Population*

Parameter 21 GW (n 5 9) 22 GW (n 5 23) 23 GW (n 5 16) 24 GW (n 5 2) 25 GW (n 5 2)

Biparietal diameter, mm 53.4 6 1.1 55.8 6 1.7 57.5 6 1.6 58.5 6 0.6 63.2 6 1.6

Abdominal circumference, mm 164.8 6 7.9 175.8 6 6.6 181.6 6 5.2 193.3 6 9.4 207.7 6 7.5

Femur length, mm 37.0 6 2.3 40.0 6 1.7 41.4 6 1.3 46.0 6 1.4 45.1 6 1.2

Skull perimeter, mm 191.1 6 4.7 199.6 6 6.3 206.1 6 5.7 215.1 6 1.6 226.8 6 4.4

Cephalic index 0.78 6 0.02 0.78 6 0.03 0.78 6 0.03 0.76 6 0.02 0.80 6 0.02

* GW 5 gestational week.

466 Cleft Palate–Craniofacial Journal, July 2011, Vol. 48 No. 4

lower gum in the fetus. Since in the fetus teeth are not

erupted yet, A9 and B9 could be considered as homologues

of the reference points A and B, which are the most

posterior points of the anterior maxillary and mandibular

contours, respectively. These landmarks are widely used in

anthropology and dentofacial orthopedics to evaluate theprojection of the maxilla and the mandible relative to the

anterior cranial base (Gugny et al., 1957; Zide et al., 1981;

Hurst et al., 2007). The last landmark, the gnathion (Gn), is

the most inferior and ventral point of the mandibular

symphysis.

The anterior cranial base is the most popular reference

line to measure the projection of the maxilla and the

mandible. Then, the antero-posterior jaw discrepancy

(A9NB9) can be measured relative to the same reference

line. For our cephalometric analysis, we chose the PN line

that has the N point as the most anterior landmark and the

P point as the most posterior landmark. Usually, the sella

turcica is used as posterior landmark, but it is difficult to

observe in ultrasound examination (Tortill, 1986; Escobar

et al., 1988; Escobar et al., 1990). Roelfsema et al. (2007)

localized the sella turcica at the intersection of the axis of

the petrous portion of the temporal bone and the edge of

the sphenoid bone between its frontal and occipital parts.

Looking at the images published in this article, it seemed to

us that the position of the sella turcica corresponded to the

ossification center of the basispostsphenoid. Thus, we

located the P point at the endocranial extremity of the

midsphenoid synchondrosis, which appears as a hypoecho-

genic dark line between the ossification center of the

basipresphenoid and the basipostsphenoid.

The projections of the maxilla and mandible were then

measured using the PNA9 and PNB9 angle, respectively.

The A9NB9 angle was calculated to evaluate the antero-

posterior jaw discrepancy. The protrusion of the chin was

calculated with the PNGn angle. The NPGn angle, which

FIGURE 1 Digitalized picture of a fetus at 33 GW showing the landmarks

used in AMSP (see text for comments).

FIGURE 2 Left: Acquisition of the UMSP. The white arrow indicates the midsphenoidal synchondrosis. Right: Measurement of the angle.

TABLE 2 Differences of the Repeated Measures in the Anatomical

and Ultrasound Populations*

Mean 1 Mean 2 F pCCC of Lin(IC 95%)

Anatomical PNA9 83.1 83.2 0.001 .97 0.95 (0.70, 0.99)

PNB9 78.2 77.4 0.09 .77 0.96 (0.80, 0.99)

PNGn 68.9 68.6 0.019 .89 0.96 (0.77, 0.99)

NPGn 69.6 69.3 0.002 .96 0.65 (20.15, 0.93)

Ultrasound PNA9 80.0 80.6 0.39 .54 0.31 (20.15, 0.66)

PNB9 71.5 72.6 0.79 .38 0.57 (0.16, 0.81)

PNGn 67.9 68.7 0.48 .49 0.51 (0.06, 0.78)

NPGn 71.9 69.7 2.9 .10 0.60 (0.23, 0.82)

* Values are expressed in degrees.

Captier et al., ULTRASOUND FETAL CEPHALOMETRIC ANALYSIS 467

corresponds to the facial angle, was used to describe the

direction of mandibular growth relative to the anterior

cranial base.

Analyses

Measurement reliability was evaluated using one-way

analysis of variance (ANOVA) for repeated measurements.

Specifically, three randomly selected fetuses from the

anatomical sample were measured twice (right and left

side) by one investigator (G.C.) in 1 week. For the

ultrasound sample, the image of the fetus head profile in

the medial sagittal plane was recorded twice (the first at the

beginning of the ultrasound examination and the second at

the end) in 17 randomly selected pregnant women by one

investigator (J.-M.F.). The reproducibly of the measure-

ments was assessed by the concordance correlation

coefficient (CCC) of Lin (1989).

The relationships between variables were first evaluated

with nonparametric methods and then the bivariate trends

were modeled by linear regression fitting. The age of the

fetuses, expressed in GW, served as the independent

FIGURE 3 Boxplot of the antero-posterior jaw discrepancy in the

anatomical (A9NB9 anat) and ultrasound (A9NB9 US) sample. GW =gestational week.

FIGURE 4 Scatter diagram and linear regression of the projection of the mandible (PNB9 values) and of the chin (PNGn values) according to the projection of

the maxilla (PNA9 values). Anatomical sample (values are expressed in degree).

468 Cleft Palate–Craniofacial Journal, July 2011, Vol. 48 No. 4

variable, and the most adapted and parsimonious models

were generated for each measurement. The level of

significance chosen was .05. The model assumption was

checked by calculating the residuals and plotting them.

RESULTS

Measurement Reliability

The ANOVA analysis of repeated measurements showed

that errors occurred but that they were negligible in

comparison to the biological variation among individual

fetuses (Table 2). In the ultrasound sample, variation was

more frequent, especially for the facial angle (NPGn). The

CCC was lower in the ultrasound sample because the

measurements were done on two different UMSP. The

concordance of the UMSP acquisition was not calculated.

Anatomical Sample

In the anatomical sample, the value of the PNA9 angle

(which describes the projection of the maxilla) increased

slightly but significantly between the second and the third

trimester (Table 3). Conversely, the projection of the

mandible (PNB9) and of the chin (PNGn) did not

significantly change between the second and third trimester.

The absolute values of A9NB9, which describe the antero-

posterior jaw discrepancy, increased between the second

and the third trimester; however, this change was not

significant (p 5 .14). The mean A9NB9 value was 5.5u 6

0.5u (n 5 18), and an important individual variation was

observed in the second trimester (Fig. 3). The projections of

the mandible (PNB9) and of the chin (PNGn) were well

determined by the projection of the maxilla (PNA9), even

though this last value increased significantly between the

two trimesters (Fig. 4). The antero-posterior jaw relation-

ship was positively correlated with PNA9 and negatively

correlated with PNB9, albeit weakly in both cases (Table 4).

The facial angle (NPGn) did not change between the

second and third trimester, and it was negatively correlated

with the PNA9, PNB9, and PNGn angles. As a conse-

quence, the direction of the facial growth relatively to the

anterior cranial base was not modified.

Ultrasound Sample

The analysis of the results (ANOVA) shows that the

measures of the different angles did not change significantly

in the different age groups (p between .35 and .65). The

values of the projection of maxilla (PNA9), of mandible

(PNB9), and of chin (PNGn) are reported in Table 5. The

mean value for the A9NB9 angle was 8.2u 6 2.7u (n 5 52),

and a huge individual variability was observed among

fetuses (Fig. 3). Conversely, no significant differences

between female and male fetuses, whatever the age, were

observed (Table 6), as reported also by Houpt (1970), who

analyzed 69 fetuses between 12 and 19 GW.

As in the anatomical study, the variations of PNA9,

PNB9, and PNGn were negatively and significantly

correlated with those of the facial angle NPGn (Table 7).

Similarly, the projection of the mandible (PNB9) and chin

(PNGn) was well determined by the projection of the

maxilla PNA9 (Fig. 5).

Finally, the ultrasound and anatomical data concerning

second-trimester fetuses were compared using the bootstrap

method. The slopes of the model of the linear regression of

PNA9 (IC [21.70, 0.46], p . .05) and A9NB9 (IC [20.05,

1.54], p . .05) on fetuses’ age were similar in both samples;

conversely, those of PNB9 (IC [22.63, 20.21], p , .05),

PNGn (IC [22.77, 20.39], p , .05), and NPGn (IC [21.65,

20.41], p , .05) were different. The covariance analysis of

the cephalometric variables with the age and the cephalo-

metric analysis type (anatomical or ultrasound) showed

that the value of the PNA9 was not different, but the value

of A9NB9 was 3.23u 6 0.78u greater in ultrasound than

anatomical cephalometric analysis (Table 8).

DISCUSSION

In this preliminary work, we show that the antero-

posterior jaw relationship could be determined accurately

by using ultrasound images of the fetus head profile in the

TABLE 3 Results of the Cephalometric Analysis in the Second and Third Trimester Anatomical Sample Population

Measurements Rho* p Regression Curve Second Trimester (mean 6 SD) Third Trimester (mean 6 SD)

PNA9 .55 ,.05 76.4 + 0.25 GW 81.8 6 3.3 83.9 6 4.2

PNB9 .25 .31 73.9 + 0.12 GW 77.1 6 4.2 76.8 6 3.8

PNGn .32 .18 65.6 + 0.12 GW 68.5 6 3.9 69.0 6 2.1

A9NB9 .28 .26 2.5 + 0.12 GW 4.7 6 2.2 7.1 6 0.8

NPGn .07 .78 65.3 + 0.09 GW 66.6 6 3.9 69.6 6 1.6

* Spearman’s coefficient; values are expressed in degrees.

TABLE 4 Bivariate Correlation Between Angles in the Anatomical

Sample Population (Spearman’s Coefficient)

PNA9 PNB9 PNGn NPGn A9NB9

PNA9 1

PNB9 .80* 1

PNGn .67* .71* 1

NPGn 2.28 2.48** 2.55** 1

A9NB9 .19 2.29 2.22 .51** 1

* p , .01.

** p , .05.

Captier et al., ULTRASOUND FETAL CEPHALOMETRIC ANALYSIS 469

medial sagittal plane taken during the routine mid-

gestation ultrasound scan.

Although there was an absolute difference between

anatomical and ultrasound measurements (3.23u), the slope

of the regression line on fetuses’ age of the antero-posterior

jaw discrepancy (A9NB9) relative to the anterior cranial

base as reference line was comparable in AMSP and UMSP

images during the fetal period. Furthermore, the antero-

posterior jaw relationship remained unchanged during the

two last trimesters, although the anterior projection of the

maxilla increased slightly with age in the anatomical

samples, as already reported by Ford (1956). The projec-

tions of the maxilla (PNA9) measured in the AMSP and

UMSP were also comparable (i.e., 81.8u versus 80.0u)during the second trimester, whereas the projections of

mandible (PNB9) and chin (PNGn) were different. This

difference could be explained by a backward rotation of the

mandible relative to the anterior cranial base in living

fetuses. Indeed, in utero, the projection of the mandible

cannot be controlled and the muscle tone at rest is different

than in formalin-fixed fetuses. Therefore, the mouth can be

slightly open, even if the lips are closed, and the mandible

can rotate backward. The more this rotation increases, the

more the facial angle (NPGn) increases. Nevertheless, as in

AMSP, the correlation of the maxilla, mandible, and chin

positions in UMSP was high.

On the other hand, the antero-posterior growth rate

between the inferior face and the midface changes during

the earlier fetal period (Ford, 1956). For example, in the

study by Levihn (1967), the value of the antero-posterior

jaw discrepancy was 15u from the 16 GW to birth, while it

was 8u between 12 GW and 16 GW. These results indicate

that during the routine second-trimester sonographic

examination, the antero-posterior jaw relationship is

relatively stable, as observed also in this study. So, in

practice, the prenatal diagnostic of retrognathism could be

done around 24 6 4.7 GW (Vettraino et al., 2003).

The antero-posterior jaw relationship had not beenstudied previously in prenatal ultrasound examinations.

Escobar et al. (1990, 1988) were the first to propose fetal

ultrasound cephalometry in the sagittal and coronal planes.

Four linear variables were studied in the sagittal plane, but

the antero-posterior jaw relationship relative to a cranial

base reference line was not analyzed. Tortil (1986)

suggested measuring the antero-posterior jaw relationship

relative to a cranial reference line (the cutaneous nasion-opisthocranion line) using the cutaneous subnasal and the

chin points, respectively. The opisthocranion-nasion-chin

(ONC) angle and the opisthocranion-nasion-subnasal

(ONS) angle informed about the respective projections of

the mandible and the maxilla relative to the cranial

reference. While angles evolved globally at the same rate

from 13 GW to 22 GW, ONS was more projected than

ONC after 22 GW. According to Tortil (1986), the antero-posterior jaw discrepancy increases in time with a more

marked retrognathism in the third trimester. These data

must be interpreted with caution because their reference

line corresponded to the total length of the skull and not

only to the anterior cranial base, and the landmarks used

were cutaneous. In ultrasound examination, retrognathism

had been appreciated also using the inferior facial angle

defined by the crossing of the reference and profile linesderived from cutaneous landmarks (Rotten et al., 2002).

Although these data have to be treated with caution, this

study clearly demonstrates that retrognathism can be

associated with normal size or micrognathic mandible.

Cephalometric analysis evaluates the bone relationship

by using standardized and reproducible specific x-ray

images. In prenatal 2D ultrasound imaging, the UMSP of

the fetal head can be considered as a standardized plane toexamine the fetal profile, the corpus callosum, and the

posterior fossa. In this plane, the lateral structures are not

observed, but they are not needed to study the antero-

posterior jaw relationship. To acquire the UMSP image

TABLE 6 Results of Fetal Cephalometry According to the Sex of

Fetuses (n = 40)*

Data PNA9 PNB9 PNGn NPGn A9NB9

Females, n 5 23 80.4 6 3.1 72.0 6 3.0 68.2 6 2.7 71.2 6 4.1 8.3 6 2.5

Males, n 5 17 79.0 6 3.1 71.4 6 4.9 66.8 6 4.6 71.9 6 3.7 7.6 6 3.4

p{ .64 .17 .61 .97 .48

* Values are expressed in degrees.

{ Mann-Whitney U test.

TABLE 5 Results of the Cephalometric Analysis in the Ultrasound Sample Population Between 21 and 25 GW*

Data 21 GW (n 5 9) 22 GW (n 5 23) 23 GW (n 5 16) 24 GW (n 5 2) 25 GW (n 5 2) Total (n 5 52)

PNA9 80.9 6 3.7 79.1 6 2.8 80.5 6 2.7 81.6 6 0.1 78.4 6 2.5 80.0 6 3.0

PNB9 72.6 6 3.9 71.7 6 4.1 71.5 6 2.9 73.5 6 1.3 68.7 6 2.4 71.8 6 3.6

PNGn 68.4 6 3.0 67.8 6 4.1 67.6 6 2.5 69.9 6 1.9 62.8 6 1.7 67.8 6 3.4

NPGn 69.9 6 3.8 70.7 6 4.2 71.6 6 3.9 72.9 6 1.4 74.9 6 2.3 71.1 6 3.9

A9NB9 8.2 6 8.9 7.4 6 2.9 9.0 6 2.5 8.0 6 1.4 9.7 6 0.1 8.2 6 2.7

* Values are expressed in degrees. GW 5 gestational week.

TABLE 7 Bivariate Correlation Between Angles in the Ultrasound

Sample Population (Spearman’s Coefficient)

Data PNA9 PNB9 PNGn NPGn A9NB9

PNA9 1 — — — —

PNB9 .60* 1 — — —

PNGn .53* .76* 1 — —

NPGn 2.45* 2.56* 2.62* 1 —

A9NB9 .20 2.55* 2.46* .27 1

* p , .01.

470 Cleft Palate–Craniofacial Journal, July 2011, Vol. 48 No. 4

needed to evaluate the anterior and posterior cranial base

and the total height of the skull and the face, the face of the

fetus has to be in front of the probe (Roelfsema et al.,

2007). Good visualization of the structures of the midface

in the UMSP is the most difficult thing to obtain in the

axial or coronal plan, and this problem increases even more

beyond 24 GW. In our study, the intrainvestigator’s

variability of the ultrasound measurements was high. This

variability could be explained in part because the UMSP

images acquired at the beginning and at the end of the

examination were slightly different even in the case of an

experienced practitioner. In future studies, it will be

necessary to measure the intrainvestigator’s and interinves-

tigator’s reproducibility of the acquisition of the UMSP.

The prenatal diagnosis of retrognathism is often subjec-

tive, and ultrasound cephalometric analysis is not consen-

sually used. To be useful in low-risk pregnancy, ultrasound

cephalometric analysis must be coupled with clinicalexamination and aesthetic judgment of the fetal profile.

For this, UMSP acquisition must be reproducible and at

least the antero-posterior jaw relationship should be

determined relative to the anterior cranial base. When this

ultrasound analysis method is performed in high-risk

pregnancies, some landmarks might be altered. Further

studies are necessary to increase the accuracy of ultrasound

fetal cephalometry analysis and to test its value for thediagnosis of the retrognathism.

Acknowledgments. The authors thank Jean Marc Gory for preparation

of the fetuses. They thank Elisabetta Andermarcher for critical reading of

the manuscript and Severine Thibault for the statistical analysis.

REFERENCES

Babcook CJ, McGahan JP, Chong BW, Nemzek WR, Salamat MS.

Evaluation of fetal midface anatomy related to facial clefts: use of US.

Radiology. 1996;201:113–118.

Broadbent B. A new X-ray technique and its application to orthodontia.

Angle Orthod. 1931;1:45.

Captier G, Faure JM, Baumler M, Bonnel F, Daures JP. Anatomy and

growth of the fetal soft palate: a cadaveric study to improve its

ultrasonographic observation. Cleft Palate Craniofac J. 2008;45:

439–445.

Chang HP. Assessment of anteroposterior jaw relationship. Am J Orthod

Dentofacial Orthop. 1987;92:117–122.

Chitty LS, Campbell S, Altman DG. Measurement of the fetal

mandible—feasibility and construction of a centile chart. Prenat

Diagn. 1993;13:749–756.

Dommergues M, Bessis R, Henrion R. Report of the French Comite

national technique de l’echographie de depistage prenatal (prenatal

FIGURE 5 Scatter diagram and linear regression of the projection of the mandible (PNB9 values) and of the chin (PNGn values) according to the projection of

the maxilla (PNA9 values). Ultrasound sample (values are expressed in degree).

TABLE 8 Covariance Analysis of the Cephalometric Variables:

Analysis Type = Anatomical or Ultrasound

Estimator Standard Deviation p

PNA9 Analysis type 21.57 0.923 .092

Age 0.23 0.09 .014

PNB9 Analysis type 24.81 1.12 ,.001

Age 0.09 0.11 .395

PNGn Analysis type 211.36 1.06 ,.001

Age 0.15 0.10 .169

NPGn Analysis type 26.42 1.17 ,.001

Age 0.19 0.12 .104

A9NB9 Analysis type 3.23 0.78 ,.001

Age 0.14 0.080 .089

Captier et al., ULTRASOUND FETAL CEPHALOMETRIC ANALYSIS 471

ultrasound): what are the practical consequences? [in French]. Gynecol

Obstet Fertil. 2006;34:1090–1095.

Escobar LF, Bixler D, Padilla LM, Weaver DD. Fetal craniofacial

morphometrics: in utero evaluation at 16 weeks’ gestation. Obstet

Gynecol. 1988;72:674–679.

Escobar L, Bixler D, Padilla L, Weaver D, Williams C. A morphometric

analysis of the fetal craniofacies by ultrasound: fetal cephalometry.

J Craniofac Gen Dev Biol. 1990;10:19–27.

Faure JM, Baumler M, Bigorre M, Captier G, Boulot P. Prenatal

diagnosis of an isolated incomplete V-shaped cleft palate using a new

three-dimensional ultrasound technique investigation. Surg Radiol

Anat. 2007;29:695–698.

Faure JM, Baumler M, Boulot P, Bigorre M, Captier G. Prenatal

assessment of the normal fetal soft palate by three-dimensional

ultrasound examination: is there an objective technique? Ultrasound

Obstet Gynecol. 2008;31:652–656.

Faure J, Captier G, Baumler M, Boulot P. Sonographic assessment of

normal fetal palate using three-dimensional imaging: a new technique.

Ultrasound Obstet Gynecol. 2007;29:159–165.

Ford E. The growth of the foetal skull. J Anat. 1956;90:63–72.

Gugny G, Delattre A, Fenart R. Definitions et determinations des points

craniens, faciaux mandibulaires et dentaires. Cahiers Odonto Stoma-

tologiques. 1957;7:34–66.

Houpt M. Growth of the craniofacial complex of the human fetus.

Am J Orthod. 1970;58:373–383.

Hurst CA, Eppley BL, Havlik RJ, Sadove AM. Surgical cephalometrics:

applications and developments. Plast Reconstr Surg. 2007;120:92e–104e.

Levihn W. A cephalometric roentgenographic cross-sectional study of the

craniofacial complex in fetuses from 12 weeks to birth. Am J Orthod.

1967;53:822–848.

Lin LI. A concordance correlation coefficient to evaluate reproducibility.

Biometrics. 1989;45:255–268.

Nicolaides KH, Salvesen DR, Snijders RJ, Gosden CM. Fetal facial

defects: associated malformations and chromosomal abnormalities.

Fetal Diagn Ther. 1993;8:1–9.

Oktay H. A comparison of ANB, WITS, AF-BF, and APDI measure-

ments. Am J Orthod Dentofacial Orthop. 1991;99:122–128.

Otto C, Platt LD. The fetal mandible measurement: an objective

determination of fetal jaw size. Ultrasound Obstet Gynecol. 1991;1:

12–17.

Paladini D, Morra T, Teodoro A, Lamberti A, Tremolaterra F, Martinelli

P. Objective diagnosis of micrognathia in the fetus: the jaw index.

Obstet Gynecol. 1999;93:382–386.

Roelfsema NM, Grijseels EW, Hop WC, Wladimiroff JW. Three-

dimensional sonography of prenatal skull base development. Ultra-

sound Obstet Gynecol. 2007;29:372–377.

Roelfsema NM, Hop WC, Wladimiroff JW. Three-dimensional sono-

graphic determination of normal fetal mandibular and maxillary size

during the second half of pregnancy. Ultrasound Obstet Gynecol.

2006;28:950–957.

Rotten D, Levaillant JM, Martinez H, Ducou le Pointe H, Vicaut E. The

fetal mandible: a 2D and 3D sonographic approach to the diagnosis of

retrognathia and micrognathia. Ultrasound Obstet Gynecol. 2002;19:

122–130.

Tortil J-M. Etude de la croissance du complexe cranio-maxillo-mandi-

bulo-rachidien chez le foetus in utero au moyen de l’echographie.

Nancy, France: Universite de Nancy I; 1986, Dissertation.

Tsai MY, Lan KC, Ou CY, Chen JH, Chang SY, Hsu TY. Assessment of

the facial features and chin development of fetuses with use of serial

three-dimensional sonography and the mandibular size monogram in a

Chinese population. Am J Obstet Gynecol. 2004;190:541–546.

Vettraino IM, Lee W, Bronsteen RA, Harper CE, Aughton D, Comstock

CH. Clinical outcome of fetuses with sonographic diagnosis of isolated

micrognathia. Obstet Gynecol. 2003;102:801–805.

Zalel Y, Gindes L, Achiron R. The fetal mandible: an in utero

sonographic evaluation between 11 and 31 weeks’ gestation. Prenat

Diagn. 2006;26:163–167.

Zide B, Graysson B, McCarthy JD. Cephalometric analysis : part I. Plast

Reconstr Surg. 1981;68:816–823.

472 Cleft Palate–Craniofacial Journal, July 2011, Vol. 48 No. 4