Cephalometric norms for craniofacial asymmetry using submental-vertical radiographs

Thomas G. Arnold, DDS, MS," Gary C. Anderson, DDS, MS," and William F. Llljemark, DDS, PhD c Minneapolis, Minn.

The submental-vertical (S-V) radiograph has become popular in the assessment of craniofacial asymmetry because of ease in identifying reliable midline reference structures. To date, no standards of asymmetry that use the S-V projection have been reported. Submental-vertical radiographs were obtained on 44 adults. Subjects were excluded if previous orthodontic treatment or temporomandibular joint symptoms were reported. With the use of a system of asymmetry analysis developed by Ritucci and Burstone, asymmetry was reported for cranial base, zygomaxillary complex, and mandibular structures. Mean and standard deviations were used to report the asymmetry values across 23 pairs of anatomic landmarks. The data showed that asymmetry is present to some degree in all landmarks and patients., Further, strong asymmetry associations existed between landmarks within patients, with most high positive correlation coefficient values found betweer) regionally or geometrically related points. A refined version of the asymmetric analysis was presented that included easily identified and clinically re!evant points. This system is more feasible for the orthodontic clinician. (AM J ORTHOt3 DENTOFAC ORTHOP 1994;106:250-6.)

The accurate diagnosis and treatment plan- ning of cases with skeletal asymmetry is important be- cause treatment options are available in which skeletal components of the craniofacial complex can be altered. Orthognathic surgery, surgical and nonsurgical maxil- lary expansion, headgear, and functional appliances are realistic and commonly used treatment modalities in which skeletal structures and craniofacial growth can be modified.' In light of these advancements, it is im- portant that the clinician accurately determine the site and assess the degree to which skeletal disharmony contributes to a given malocclusion.

Several systems and methods for assessing skele- tal asymmetry have been reported. 2~ Radiographic analyses s have been used extensively because of the ease and accuracy in measuring the underlying sup- portive structures. The posterior-anterior (P-A) projection 69 has been widely used since horizontal and vertical relationships can be delineated clearly. Diffi- culty in identifying landmarks suitable for a midsagittal reference system with the P-A radiograph led to the use of the submental-vertical (S-V) projection. 10 Patient po- sitioning for the S-V projection, with the Frankfort hor-

From the University of Minnesota. 'Assistant Clinical Professor, Department of Preventive Sciences, School of Dentistry. and in private practice of orthodontics, Lakeville; Minn. bAssociate Professor, Department of Restorative Sciences, School of Dentistry. 'Professor and Chair. Department of Diagnostic and Surgical Sciences, School of Dentistry. Copyright 9 1994 by the American Association of Orthodontists. 0889-5406/94/$3.00 + 0 811142489

250

izontal plane parallel with the film cassette, projects the facial bones upward to allow for clear visualization of the anatomic structures of the cranial base.

By using cranial base landmarks to develop a mid- line reference system, Ritucci and Burstone" described a system for assessment of craniofacial asymmetry with the S-V projection. In this system, asymmetry of the cranial base, zygomaxillary complex, and mandible could be assessed. Measurements to assess bilateral symmetry within each component of the skull were made relative to a coordinate axis system developed from anatomic points within that skull component. All anatomic landmarks were measured relative to a co- ordinate system developed from foramen spinosa, the most reproducible cranial base landmark.

At the present time no cephalometric standards exist to report the normal range of asymmetry within the general population with a S-V radiograph. Without a clinical "yardstick" for individual evaluation, the use- fulness of the S-V radiograph to the clinician is limited. The objectives of this study were to develop and report cephalometric norms of asymmetry for the Ritucci and Burstone system of analyzing submental-vertical radio- graphs, and to develop a simplified system for clinical use based on these norms.

MATERIALS AND METHODS

Submental-vertical radiographs were obtained from 44 adult white dental students from the University of Minnesota School of Dentistry. Subjects ranged in,age from 18 to 34 years, with a mean age of 25.4 years. Twenty-eight men and

American Journal of Orthodontics and Dentofaciat Orthopedics Amold~ Anderson, and Liljemark 251 Vohtme 106, No. 3

16 women participated in the study. Subjects were excluded from participation in the study if they presented with a positive history to any of the following: (1) any type of orthodontic treatment, (2)jaw fractures or jaw surgery, (3) temporoman- dibular joint (TMJ) dysfunction and/or pain symptoms, (4) removable or fixed dental prosthesis, and (5) missing teeth mesial to permanent second molars.

The S-V projection was made in a Wehmer cephalostat (Wehmer Corp., Franklin Park, I11.) by rotating the patient's head posteriorly until the Frankfort horizontal plane was par- allel with the film cassette. The head position was fixed by positioning the nasal assembly beneath the chin. The cathode- to~earrod distance was a standard 60 inches, and the earrod- to-film distance was fixed at 16 cm. Mandibles of all patients were positioned in the intercuspal position (IP) before rotating the head posteriorly.

A submental-vertical cephalometric analysis, proposed by Ritucci and Burstone," was used to assess asymmetry in the cranial base, zygomaxillary complex, and mandible. This system uses bilateral anatomic points to determine reference coordinate systems to which other landmarks are related. The symmetry of 22 paired and unpaired anatomic points was evaluated through comparison with these reference systems (Fig. 1). For paired structures, the distance to the midline was determined for both landmarks and the difference in horizontal distance was calculated. For unpaired points, the horizontal distance to the midline was determined.

The Ritucci-Burstone system analyzes structures within the cranial base, zygomaxillary complex, and mandible. The landmarks involved in the cranial base assessment are as follows: (1) foramina spinosa points (FS); (2) posterior cranial vault points (PCV); (3) middle cranial fossa points (MCF); (4) basion (Ba), and (5) opisthion (Op). Measure- ments of asymmetry within the cranial base structures were made relative to a coordinate system developed from right and left foramen spinosa. The interspinosal line served as the x-axis, and the interspinosal axis served as the y-axis.

Zygomaxillary complex structures involved are as fol- lows: (1) pterygomaxillary fissure (PTM); (2) buceale (Be); (3) zygion points (ZP); (4) anterior cranial vault points (ACV); (5) angulare points (A); (6) anterior vomer points (AVP); (7) posterior vomer points (PVP); and (8) maxillary dental mid- line (MxDM). A coordinate system developed from the pter- ygomaxillary line and its perpendicular bisector was used to measure asymmetry within the zytomaxillary complex..

The anatomic points tested in the mandible area are as follows: (1) gonion points (Go); (2) condylion anterioris (CA); (3) condylion posterioris (CP); (4) condylion lateris (CL); (5) condylion medialis (CM); (6) coronoid process points (CPP); (7) first molar points (FMP); (8) gonial geometric midline of the mandible (GGM); (9) condylar geometric midline of the mandible (CGM); and (10) mandibular dental midline (MnDM). A coordinate system developed from the condylion line and its perpendicular bisector was used to determine mandibular asymmetry.

Pearson's correlation coefficients '2 were calculated to as- sess the relationships between anatomic points within specific patients. Each pair of landmarks was assessed to determine

9 Idx l3b4

FMP_ Mn l )~~,~ FMP

CPP , ~ ~CPP

Fig. 1. Structures visible in submental-vertical projection: MxDM, maxillary dental midline; MnDM, mandibular dental mid- line; AVP, anterior vomer point; PVP, posterior vomer point; FMP, first molar point; A, angulare; Bc, buccale; MCF, middle cranial fossa; PTM, pterygomaxillary fissure; CPP, coronoid pro- cess point; ZP, zygion point; ACV, anterior cranial vault; FS, foramen spinosum; C, condyle; Go, gonion; Ba, basion; Op, opisthion.

associations between the two related points, i.e., if One land- mark exhibited asymmetry in a certain direction, how were other landmark asymmetry values affected. Correlation co- efficients for all pairs of landmarks were calculated across the sample population relative to foramen spinosum, ptery- gomaxillary fissure, and condylion anterioris reference systems.

Actual values of asymmetry, as opposted to absolute val- ues, were analyzed to determine any right or left dominance patterns within any landmarks across the study group. Paired t tests were used to determine whether mean values differed significantly from zero.

Fifteen radiographs were selected at random to determine tracing and measurement error. Ten of these radiographs were traced a second time to assess landmark identification error. Measurement error for each landmark was determined by calculating the standard deviation of the repeated events. The standard deviation values were pooled, and a standard devia- tion for each landmark across the subject group was calcu- lated. Five acetate tracings were selected to analyze error in the process of data entry. Digitizing error was calculated with the standard deviation and pooled standard deviation of the repeated processes, as described previously.

A refinement of the Ritucci-Burstone system was devel- oped for use clinically. The scaled down version included those points that were reliably identified and could be affected by clinical treatment modalities. Related points as determined by correlation coefficients were represented in the clinical system by the one most reliable landmark within the group, as indicated by the lowest measurement error value.

252 Arnold, Anderson, and Liljernark American Journal of Orthodontics and Dentofacial Orthopedics September 1994

10-

9 .

8 .

5

r 3

:~ 2

I

O,

Landmark

Fig. 2. Mean and one standard deviation values calculated from absolute difference values for land- marks related to foramen spinosUm reference system.

10

i 6- 5- 4"

e- 3-

:~ 2

' . . . .

Landmark

Fio. 3. Mean and one standard deviation values for zygomax- illary complex landmarks related to pterygomaxillary fissure ref- erence system.

RESULTS

Fig. 2 shows mean and standard'deviation values calculated from absolute values of right minus left dif- ferences in distances frorn' the determined centerline. Unpaired t tests on male and female subgroups resulted in no statistically significant difference between sexes.

Fig. 3 shows mean and standard deviation values for zyg0maxillary' anatomic points related to the' pterygo- maxillary reference system. No statistically significant Sex differences were noted for zygomaxillary land- marks.

Fig. 4 shows mean and standard deviation values for mandibular landmarks compared to the condylion an- terioris reference system. No statistically significant dif- ferences resulted between male and female subgroups.

9

8

7- E

, - 3

~ 2

1

:E

Landmark

Fig. 4. Mean and one standard deviation values for mandibular landmarks related to condyl[on anterioris reference system.

Correlation coefficients, calculated on fight minus left differences for al! pairs ofiandmarks, ranged'from -0 .33 (ACV and cp) to 0.98 (CA and CP) with the foramen spinosum reference S),stem. Coefficient val/ies greater than 0.75 were reported for ACV and PCV,'A and Be, ZP and A, AVP and PVP, AVP and MxDM, AVP and MnDM, PvP and MxDM, PVP and MnDM, MxDM and FMP, MxDM and MnDM, MxDM ~md GGM, Go and CA, Go and CP, CA and CP, and FMP and MnDM. Correlation coefficients Were also' calcu- lated for intramaxillary and intramandibular landmarks relative to within region reference systems. Maxillary values ranged from -0.41 (ACV and MxDM) to 0.79 (AVP and PVP). Coefficient values 0.75 or greater were noted for AVP and PVP, ACV arid ZP. Mandibular val- ues ranged from -0 .58 (CM and FMP) to 0.93 (FMP and MnDM). Values 0.75 or greater were noted for FMP and MnDM.

American Journal of Orthodontics and Dentofacial Orthopedics Arnold, Anderson, attd Liljemark 253 Volume 106, No. 3

8 7s 6: a- 4-

i. T .... IT 2

Landmark

Fig. 5. Mean and one standard deviation values from actual right minus left difference values for all landmarks using foramen spinosum reference system.

Fig. 5 shows mean and standard deviation values calculated from actual right minus left differences with the foramen spinosum reference system. Mean values ranged from -2 .33 mm for ACV to 2.79 mm for first molar point. Standard deviation values ranged from 1.40 mm for PVP to 5.50 mm for ACV. Paired t tests revealed mean values statistically different from zero at the p < 0.05 significance level for Op, Ba, PTM, ACV, MxDM, Go, Ca, Cp, FMP, MnDM, and GGM with the foramen spinosum reference system.

Figs. 6 and 7 show absolute difference values for zygomaxillary and mandibular structures with their re- spective reference systems. Mean values for zygomax- illary structures ranged from -3 .00 mm for ACV to 0.22 mm for MxDM. Mean values for mandibular structures ranged from - 1.44 mm for CM to 1.83 for FMP.

Results obtained for the 10 duplicate acetate tracings relative to the foramen spinosum reference system ex- hibited standard deviation values ranging from 0.00 mm to 3.42 ram. Standard deviations were pooled to assess intrasubject and intralandmark error. The pooled stan- dard deviation value for all landmarks across the 10 subjects was 0.88 mm. Standard deviation values for pterygomaxillary fissure landmarks ranged from 0.00 to 2.48 mm with a pooled standard deviation of 0.87 nun, and from 0.00 to 3.03 mm with a pooled value of 0.90 mm for mandibular structures.

Five acetate tracings were used to analyze error in the data entry process. Standard deviations of right mi- nus left differences for the two digitizing processes with the foramen spinosum reference system ranged from 0.00 to 0.69 mm. The pooled standard deviation value

for all landmarks across the five subjects was 0.29 mm. Values obtained from the pterygomaxillary fissure ref- erence system ranged from 0.01 to 1.05 mm with a pooled value of 0.38 ram, and from 0.01 to 0.53 mm with a polled value of 0.20 mm for mandibular struc- tures.

DISCUSSION

Mean difference values reported in this study reveal that certain landmarks exhibit more asymmetry than others. Mean values for differences within the three regions (cranial base, zygomaxillary, and mandibular) did not differ considerably. Vig and Hewitt ~3 reported greatest overall symmetry in the dentoalveolar regions, concluding that compensatory changes occur with!n the dentoalveolar regions during growth and development enabling symmetric function and maximum intercup- sation to occur despite underlying asymmetries within the jaws. Although minimal data are obtained from landmarks within the dentoalveolar regions in the Ri- tucci-Burstone system, the fact that maxillary and man- dibular dental midlines were so closely related, as shown by a correlation coefficient of 0.93, would seem to support the theory of Vig and Hewitt.

Correlation coefficients were calculated to assess the relationships between anatomic points within spe- cific patients. All pairs of landmarks were used to assess relationships of points across the subject group. Al- though the majority of coefficients were positive, neg- ative values were noted for several pairs of landmarks. Negative correlation coefficient values can be explained by deflection or proliferation of adjacent structures as a result of asymmetric changes. Rotational changes, as

254 Arnold, Anderson, and Liljemark American Journal of Orthodontics and Dentofacial Orthopedics September 1994

=

==

I " 0'"

.! :

-2 ":

.4 - -5 * -6 - -7 ": s :

||=,,

Landmark

i:i9. 6. Mean and one standard deviation values from actual right minus left difference values for 7ygomaxillary comple landmarks using pterygomaxillary reference system.

7-

4: 3: 2: ,i 0

-2

-3

-4

-5 .6

-7

-8

T -r

F I

I i

Landmark

Fig. 7. Mean and one standard deviation values from actual right minus left difference values for mandibular landmarks us- ing condylion anterioris reference system.

in the case of glenoid fossa remodeling and its resultant change in mandibular position, could also explain neg- ative values.

Absolute values of differences between right and left horizontal distances to the midline plane were used to determine the norms for each landmark. This allowed for an assessment of average difference values between right and left landmarks, eliminating the possibility of positive and negative difference values cancelling each other in mean calculations. A determination of right or left dominance across the 44 subjects was performed by evaluating the actual difference values. Paired t tests indicated that the majority of landmarks (22 of 38) analyzed in this investigation did exhibit mean values

that differed significantly from zero. Although statis- tically significant differences were exhibited, the clin- ical significance of the actual difference values would appear minimal, considering the mean values (all

American Journal of Orthodontics and Dentofacial Orthopedics Artlold, Anderson, and ~..j rt- 255 Volume 106, No. 3

gions of landmarks had differences in error measure- ment involved in their identification. The greatest errors in identification in this study were reported in the cranial base landmarks, exemplified by MCF, PCV, Ba, and Op being four of the most variable landmarks. Zygo- maxillary and mandibular landmarks exhibited approx- imately equal amounts of error in identification. These regional differences can be explained by the investi- gator's noted difficulty in locating the specific cranial base landmarks in certain radiographs.

The duplicate entry of data points reveals that min- imal (

256 Arnold, Anderson, and Liljemark American Journal of Orthodontics and Dentofacial Orthopedics September 1994

Table I. Mean and one standard deviation values for landmarks in simplified system for clinical use

Landmark [ Mean(ram) [ StDev

PTM 2.43 1.90 PVP 1.11 0.85 AVP 1.76 1.43 MxDM 2.16 1.69 CGM 1.23 1.10 GGM 1.69 1.52 MnDM 2.19 1.52 CP 2.75 2.17 CA 2.61 2.13

orthodontically in this patient may best be achieved by combining maximum dental movement in the mandible with orthopedic manipulation in the maxilla.

SUMMARY AND CONCLUSIONS

Cephalometric standards were reported on S-V ra- diographs obtained from 44 adult subjects. Correlation coefficients were obtained on all combinations of land- marks to assess the relationships and tendencies noted between regions and specific points. An evaluation was performed on error involved in landmark identification and data entry procedures. From the developed sub- mental-vertical norms, a refined system suitable for clinical use was proposed.

The following conclusions were made: 1. Asymmetry was present to some degree in all

patients and landmarks. 2. No sex differences were noted for asymmetry

across the sample population. 3. Large values of asymmetry were present in each

landmark across the subject group, and most patients displayed asymmetry approaching 2 standard deviation units for at least one anatomic point.

4. Strong asymmetry associations existed between landmarks within patients, with most high pos- itive correlation coefficient values found be- tween regionally or geometrically related points.

5. Error measurements revealed that not all land- marks were identified with equal accuracy, and that the majority of the total error was associated with landmark identification.

REFERENCES 1. Proffit W. Contemporary orthodontics. St. Louis: CV Mosby,

1986. 2. Sved A. An analysis of the most important diagnostic methods

used in orthodontia. Angle Orthod 1931;!:139-60. 3. Cheney EA. Dentofacial asymmetries and their clinical signifi-

cance. AM J ORa'ROD 1961;47:814-29. 4. Sutton PRN. Lateral facial asymmetry--methods of assessment.

Angle Orthod 1968;38:82-92. 5. Broadbent BH. A new x-ray technique and its application to

orthdontia. Angle Orthod 1931;1:45-66. 6. Harvold E. Cleft lip and palate. AM J OR'rHOD 1954;44:493-506. 7. Mulick JF. An investigation of craniofacial asymmetry using the

serial twin-study method. AM J OR'roOD 1965;51:112-29. 8. Vig PS, Hewitt AB. Asynmmetry of the human facial skeleton.

Angle Orthod 1975;45:125-9. 9. Grummons DC, Kappeyne Van de Coppello MA. A frontal asym-

metry analysis. J Clin Orthod 1987;21(7):448-65. 10. Berger H. Problems and promises of basilar view cephalom-

grams. Angle Orthod 1961;31:237-45. I 1. Ritucci R, Burstone CJ. Use of the submental vertical radiograph

in the assessment of asymmetry. Thesis submitted in partial ful- fillment of the requirements for certificate in orthodontics. Far- mington: University of Connecticut, 1981.

12. Colton T. Statistics in medicine. Boston: Little, Brown and Com- pany, 1974.

13. Vig PS, Hewitt AB. Is craniofacial asymmetry and adaptation for masticatory function an evolutionary process.'/ Nature 1974;248:165.

14. Truque MG. A study of the asymmetries of the maxillary dental arch in relation to raphe palatinus and the mediastinal plane of the upper facial skeleton. [Thesis.] Ann Arbor: University of Michigan, 1956.

15. Mayers CA. Asymmetry of the maxillary and mandibular den- tition with respect to the palatine raphe. Thesis submitted in partial fulfillment of the requirements for the Degree of Master of Science. Cleveland, Ohio: Case Western Reserve University, 1977.

16. Forsberg CT, Burstone CJ, Hanley KJ. Diagnosis and treatment planning of skeletal asymmetry with a submental-vertical radio- graph. AM J OR'I'HOD 1984;85:224-37.

17. Baumrind S, Frantz RC. The reliability of head film measure- ments, h Landmark identification. AM J Oga3toD 1971;60:111- 27.

Reprint requests to: Dr. Thomas G. Arnold 17599 Kenwood Trail, Suite 4 Lakeville, MN 55044