Rapid maxillary expansion. A study of thelong-term effects

Paloma Velazquez, MD, DDS," Elena Benito, MD, DDS,b andLuis Alberto Bravo, MD, DDS, MS, PhDb

Alicante and Murcia, Spain

A study was made to determine whether skeletal alterations usually produced by rapid maxillaryexpansion may be compensated for in time by growth and/or comprehensive orthodontic treatment.In 30 patients, orthodontic treatment was started with rapid maxillary expansion, followed by fixedappliances, not combined with any other form of orthopedic device. Mean treatment time was 3.1years. Nine measurements from the Ricketts analysis were studied, based on lateral cephalometricradiographs. Records were taken before orthodontic treatment and after completion of activetherapy. A statistical analysis was made of the nine variables used, reflecting the vertical andanteroposterior skeletal proportions of the face, contrasting the changes before and after treatment.Of all the variables studied, the four that change with age according to the Ricketts analysis(mandibular plane angle, maxillary height, facial depth and facial convexity), yielded statisticallysignificant differences after treatment, indicative of normal growth. The five remaining variables thatremain constant with age according to the Ricketts analysis (facial axis, lower facial height, totalfacial height, palatal plane inclination and maxillary depth) showed no significant changes aftertreatment, also indicative of normal growth. (AM J ORTHODDENTOFACORTHOP1996;109:361-7.)

Studies of skeletal changes after rapid max-illary expansion (RME) have to date been carriedout using cephalometric data collected before andimmediately after active expansion. It is widely ac-cepted that these changes consist of a drop of themaxilla (generally in association with forward dis-placement) and a posterorotation of the mandible,due to the extrusion of the upper molars and theoutward inclination of the upper alveolar process.This would generally improve patients with skeletalClass III malocclusions and worsen Class Il andopen bite problems.':"

We are of the opinion that in subsequent periodsthese alterations might undergo almost completecompensation relapse. However, none of the authorsreviewed performed studies after a period that fol-lowed the interruption of RME. Only Wertz" pub-lished a study of 60 patients in which a third set ofrecords was taken. Thus, in addition to the measure-ments performed before and after active RME, herepeated the recordings after 3 months of retentionwith the passive RME device. This third set of mea-surements revealed the beginning of the return ofboth maxilla and mandible to their original positionswith maintenance of the arch length.

'Orthodontist in private practice, Alicante, Spain.'Professor of Orthodontics, University of Murcia, Spain.Copyright © 1996 by the American Association of Orthodontists.0889-5406/96/$5.00 + 0 8/1/57600

Our clinical impression is that the unfavorablechanges induced by the RME device, e.g., open biteor mandibular posterorotation, are reversible. Ingeneral, when terminating comprehensive ortho-dontic treatment we found that these undesirableeffects had been almost completely resolved. Thusgrowth, the effect of the orthodontic treatmentitself, or both compensate all the vertical alter-ations produced up to that point by the RMEdevice. Therefore the aim of the present study wasto investigate the long-term consequences of thistype of treatment on the facial skeletal pattern. Tothis effect, we evaluated 30 patients subjected toRME, final records being taken on completion oforthodontic treatment rather than immediately af-ter active expansion.

MATERIAL AND METHODS

Thirty patients (22 girls and 8 boys) from threeprivate orthodontic practices were studied. Mean patientage was 12.1 years (range 7 to 17 years) at the start oftreatment (Table I). According to Ricketts classifica-tion," the patients fell into three predominant facialtypes: dolichofacial (n = 15), mesofacial (n = 8), andbrachifacial.(n = 7). As to the type of skeletal malocclu-sion present, 15 patients were Class I, 5 Class 11,and 10Class Ill. All patients were initially treated by rapidexpansion of the upper jaw, followed by the applicationof fixed orthodontic appliances. None wore any othertype of orthopedic device. Intermaxillary elastics were

361

362 Velazquez; Benito, and Bravo American Journal of Orthodontics and Dentofacial OrthopedicsApril 1996

Fig. 1. Haas-type rapid maxillary expansion device used in all cases in this study, in positionwithin mouth.

used in 18 cases, whereas extractions were performed in15 patients. Mean treatment duration was 3.1 years.

Rapid maxillary expansion was carried out with aHaas-type apparatus (Fig. 1), consisting of four bands.The posterior bands were fitted on the first two uppermolars, and the anterior bands on the first two upperpremolars or (in the case of mixed dentition) on the firsttwo temporary molars. The anterior and posterior bandsof each hemiarch were welded together by 0.045-inchwires, which in turn were welded to a central 14 mmexpansion Hyrax-type screw. Each full turn of the screwproduced an expansion of 0.9 mm. The entire device wascovered with autopolymerizing acrylic occupying thepalatal zone.

The device activation protocol involved a quarterturn every 12 hours, starting with the teeth in crossbite,until the upper palatal cusp contacted the lower vestib-ular cusps (Fig. 2). The mean activation period was 21days, with weekly patient follow-ups. After screw liga-tion, the RME device was left in place as retention for 4months, during which conventional fixed orthodontictreatment was initiated.

All patients had complete diagnostic records beforeand after full orthodontic treatment, including lateralcephalometric radiographs. Sixty radiographs were stud-ied with measurements taken from the Ricketts cepha-lometric analysis." The following nine variables wereevaluated before and after treatment (Figs. 3, 4, and 5):

1. Variables measuring vertical skeletal proportions

• Facial axis indicates the direction of mandibulargrowth with respect to the cranial base. Norm= 90° ± 3° (mean ± SD). No change with age.

• Lower facial height relates the vertical position ofthe mandibular body to the upper jaw. Norm =47° ± 4° (mean ± SD). No change with age.

Table I. Age distribution (in years)

I ,i"i</"i SD I--x-F-l-'na-l-S-D--

Age 12.1 2.72.2 15.2

• Mandibular plane angle positions the mandibularplane with respect to Frankfort plane. Norm =26° ± 4° (mean ± SD) for 9-year-old patients, witha decrease of 1° every 3 years.

• Maxillary height indicates the vertical position ofthe maxilla with respect to the cranial base.Norm = 53° ± 3° (mean ± SD) for 9-year-old pa-tients and increases 0.5° yearly.

• Total facial height describes the position of themandibular body with respect to the cranial base.Norm = 60° ± 3° (mean ± SD). No change withage.

• Palatal plane inclination indicates the inclina-tion of the palatal plane with respect to Frank-fort plane. Norm = 1° ± 3S (mean ± SD). Nochange with age.

2. Variables measuring anteroposterior skeletal proportions

• Maxillary depth indicates the anteroposterior posi-tion of the maxilla. Norm = 90° ± 3° (mean ±SD). No change with age.

• Facial depth indicates the anteroposterior positionof the mandible. Norm = 87° ± 3° (mean ± SD)for 9-year-old patients with an increase of 1° every3 years.

• Facial convexity indicates the anteroposterior po-sition of the maxilla with respect to the anteropos-

American Joumal of Orthodontics and Dentofacial OrthopedicsVolume J09, No. 4

Veldzquez, Benito, and Bravo 363

Fig. 2. Palatal cusps of upper molars in contact with vestibular cusps of lower molars. Rapid maxillaryexpansion is terminated at this point.

Fig. 3. Four of cephalometric variables used to evaluatevertical skeletal proportions: total facial height (TFH), lowerfacial height (LFH), maxillary height (MH), and mandibularplane (MP). Arrows show increase or decrease of changesobserved after treatment.

terior position of the mandible. Norm = 2 ±2 mm (mean ± SD) for 9-year-old patients with adecrease of 1 mm every 3 years.

All radiographs were directly digitized by the same

Fig. 4. Two additional variables used to evaluate verticalskeletal proportions: facial axis (FA) and palatal plane (PP).

person, without previous tracing, and the variables mea-sured by a computerized cephalometric program (Quick-Ceph from Orthodontic Processing, Chula Vista, Calif.).Data were statistically processed by the Stat- View pro-gram (Abacus Concepts, Berkeley, Calif.). The signifi-cance of the changes observed was determined by apply-

364 Veldzquez, Benito, and Bravo

Fig. 5. Three cephalometric variables used to evaluate antero-posterior skeletal proportions: maxillary depth (MD), facialdepth (FD), and facial convexity (Fe). Arrows show increase ordecrease of changes observed after treatment.

ing the paired t test, and accepted as statistically signif-icant when p s; 0.05.

Before measurement recording and analysis of thedata, the error of the method was established. To quan-tify the error because of the localization of the cephalo-metric points involved in the measurement of the vari-ables of our study, the same operator digitized (J-monthinterval) all points used on the 60 radiographs on twoseparate occasions.

Correlation analysis revealed a mean correlation co-efficient (R) of 0.85 (SD = 0.11; range between 0.62 and0.94) for the double measurements. The mean magni-tude of error for each cephalometric variable (Xi) wasdetermined from the following mathematical formula:V~(XA -XB)2/2N.15 This magnitude was 0.57 mm in thecase of the only linear measurement used (facial convex-ity), and always under 2.0° for the angular measurements;actually, the mean value of the method error for angularmeasurements was 1.27° (SD = 0.31°; range between0.94° and 1.76°).

Finally, to evaluate separately the changes observedafter treatment in the two extreme facial types andbecause of the small size of the subgroups (brachifacialn = 7, and dolichofacial n = 15), we used the WilcoxonSigned Rank test, a nonparametric test analogous to thepaired t test.

American Journal of Orthodontics and Dentofacial OrthopedicsApril 1996

RESULTS

Table 11 shows descriptive statistical data(mean ± SD) for all the variables used, before anytreatment and after the orthodontic treatment wascompleted.

With respect to the variables that do not varywith age in the Ricketts analysis, we can see inTable 11 that the facial axis underwent no signifi-cant changes on average, which suggests that thedirection of growth remained constant. The lowerfacial height and the total facial height likewisefailed to modify on average after treatment. Thepalatal plane inclination did not vary significantlyon average with respect to Frankfurt plane, either,and the maxillary depth underwent no significantchanges on average, suggesting the absence of an-teroposterior upper jaw displacement.

On the other hand, with respect to the variablesthat do vary with age in the Ricketts analysis, wecan also see in Table 11 that the mandibular planeangle underwent a mean significant decrease of1.2°, from an initial average of 27.2° to a final 26.0°.This suggests a forward and upward mandibularrotation, i.e., in the normal direction of growth.Maxillary height in turn increased significantly 1.7°on average, reflecting a descent of the upper jawwith respect to the cranial base. i.e., in the normaldirection of growth also. Finally, facial depth in-creased 1.0° on average, indicating an anteriordisplacement of the mandible with respect to na-sion, and facial convexity decreased 1.0 mm onaverage, i.e., the maxilla receded anteroposteriorlywith respect to the mandible. Both changes were inthe normal direction of growth.

Table III shows the comparison of the cepha-lometric variables studied, before and after treat-ment, in the 15 patients with dolichofacial pattern.In general, results agree with those of the wholesample, although the changes show a lower level ofstatistical significance, probably because of thesmall number of cases.

Similar results for the same variables are shownin Table IV for the brachifacial patients (n = 7).Nevertheless, none of the changes observed werestatistically significant after treatment. This couldalso be due to the small size of this subsample.

DISCUSSION

Earlier investigations+'v":" reported a greaternumber of changes of different signs. However,these studies were always carried out on complet-ing the opening of the midpalatal suture, or nolater than 3 months after passive retention with theRME device.'

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 109, No. 4

,

Veldzquez, Benito, and Bravo 365

Table 11.Comparison of certain sample morphologic characteristics before (T1) and aftertreatment (T2)

T, T2

Proportions i SD i SD Paired t test

VerticalFacial axis (degrees) 87.5 5.1 86.9 4.3 nsLower facial height (degrees) 47.2 5.2 47.4 5.3 nsMandibular plane (degrees) 27.2 5.6 26.0 5.7 **Maxillary height (degrees) 60.8 3.8 62.5 4.0 **Total facial height (degrees) 61.7 5.9 61.6 5.3 nsPalatal plane (degrees) 1.0 5.4 1.7 4.5 ns

Anterior-posteriorMaxillary depth (degrees) 89.4 3.3 89.4 3.4 nsFacial depth (degrees) 88.7 3.8 89.7 3.6 **Facial convexity (mm) 0.7 2.7 -0.3 3.0 **

**(p :5 0.01); ns, statistically not significant.

Table Ill. Comparison of certain sample morphologic characteristics before (T1) and after treatment (T2)in the dolicholfacial group (n = 15)

T, T2 Wilcoxon

I Isigned-rank

Proportions i SD i SD test

VerticalFacial axis (degrees) 84.1 2.9 83.9 2.8 nsLower facial height (degrees) 50.1 3.7 50.0 4.5 nsMandibular plane (degrees) 30.4 4.1 29.1 4.9 nsMaxillary height (degrees) 62.1 2.9 63.9 3.4

Total facial height (degrees) 65.6 4.4 65.3 3.9 nsPalatal plane (degrees) 0.9 4.2 0.9 4.5 ns

Anterior-posteriorMaxillary depth (degrees) 89.1 3.1 88.7 3.7 nsFacial depth (degrees) 87.2 2.7 88.2 3.2Facial convexity (mm) 1.9 2.2 2.7 0.7

*(p :5 0.05); ns, statistically not significant.

According to some aurhors.Y" the facial axisopens to yield an increase in vertical facial dimen-sion shortly after finalizing RME. These and mostother authors':":" caution against the use of suchtreatment in patients with a tendency to verticalgrowth and skeletal open bites, However, in ourstudy and 3 years after concluding RME, the facialaxis was found to not vary significantly on average,This means that the mean direction of growth staysconstant, as reflected by comparing radiographsbefore and after full orthodontic treatment. Animportant consideration here is that the meanchange for the facial axis in the 15 dolichofacialsubjects studied also did not show statistically sig-nificant differences after treatment.

Lower facial height likewise failed to yield sig-

nificant changes on average after treatment, as innormal growth. It may thus be stated that, in oursample, the mandibular body axis does not vary itsposition vertically with respect to the upper jaw.

As in normal growth, the mandibular planeinclination was found to decrease on average aftertreatment. According to Ricketts, the norm for thisvalue is 26 ± 40 (mean ± sd) for 9-year-old pa-tients, and decreases 10 'every 3 years. Thus growthleads to mandibular anterorotation, due to greatergrowth posteriorly than in the anterior part of themandible. The mandibular plane angle thereforebecomes more horizontal. Authors who have in-vestigated mandibular changes immediately afterRME2,3,5,9,13 describe mandibular posterorotation.Although these authors did not employ Ricketts'

366 Velazquez; Benito, and Bravo American Journal of Orthodontics and Dentofacial OrthopedicsApril 1996

Table IV. Comparison of certain sample morphologic characteristics before (T 1) and after treatment (T2)in the brachifacial group (n = 7)

T} T2 Wilcoxonsigned-rank

i SD i SD test

VerticalFacial axis (degrees) 93.2 3.8 9l.8 3.2 nsLower facial height (degrees) 40.4 3.3 40.9 3.2 nsMandibular plane (degrees) 20.6 2.7 19.3 2.8 nsMaxillary height (degrees) 59.6 3.0 60.6 3.4 nsTotal facial height (degrees) 54.6 3.4 54.7 2.5 nsPalatal plane (degrees) -0.9 5.8 0.5 4.1 ns

Anterior-posteriorMaxillary depth (degrees) 89.6 3.4 89.7 3.3 nsFacial depth (degrees) 9l.8 3.0 92.6 2.7 nsFacial convexity (mm) -2.0 2.6 -3.1 3.0 ns

ns, statistically not significant.

cephalometric method, such movements reflect anopening of the mandibular plane angle; this iscontrary to our own observations on studying thechanges at the end of an average of 3 years of fullorthodontic treatment.

Three months after retention, Wertz" describeda systematic recovery of mandibular posterorota-tion. Costa" and Bishara and Staley? reported apartial recovery of this movement also during re-tention. Specifically, Bishara and Staley" observed a10% to 15% recovery of the skeletal changes ingeneral. It thus seems that if this movement beginsto recover or relapse in retention with the passiveRME device, posterorotation fully recovers afterremoval of the latter (and continuing full fixedorthodontic treatment for an average of 3 years).Indeed, the mandibular plane angle even closes,becoming more horizontal than at the start oftreatment. The treatment can therefore not be saidto have influenced the normal evolution of themandibular plane, for after 3 years it is seen to havechanged in the same way and almost with the sameintensity as in individuals not subjected to ortho-dontic treatment. According to this study, RMEcould be performed whenever required to secure afavorable development of the upper jaw undercompression, even when the mandibular plane isquite open.

Maxillary height was found to increase 1.70 onaverage with treatment, i.e., the upper jaw de-scended with respect to the cranial base to thatamount. This agrees with normal growth, for accord-ing to Ricketts, the angle increased OSyearly. Thus,the net increase after an average of 3 years of treat-ment would be 1.SO. The literature on vertical skel-

etal displacement of the upper jaw after RMEagrees that the latter descends, either parallel orrotating anteriorly or posteriorly, and that a 1 to2 mm descent of the upper jaw is the norm.' Thechange observed in our sample after 3 years of treat-ment is perfectly compatible with normal growth.

As to total facial height, this angle measures therelationship between the mandibular body and thecranial base. In the present study, this relationshipshowed no significant changes on average, as ithappens in normal growth.

The palatal plane inclination showed no signif-icant changes on average in the present study.According to Ricketts, this variable is measuredwith respect to the Frankfort plane. Ail authorsconsulted found the palatal plane to descend underthe effect of RME. This descent is either parallel toits initial position (as reported by Wertz" in anumber of cases) or the result of posterior rota-tion.3,4,12.13 These authors also described anteriorrotations in a smaller percentage of patients. Nev-ertheless, Sarver and Johnston'" found their di-rectly cemented device to always produce posteriorrotation. In our study and 3 years after the firstrecordings, no significant changes were observed,i.e., although the maxilla descended and, on aver-age, the palatal plane changed its inclination in 0.70

of anterior rotation (1.00 vs 1.70), this observationcannot be accepted as significant statistically.

In our sample, maxillary depth underwent nosignificant changes as with normal growth. Thus nochanges seem to have occurred in the anteroposte-rior position of the upper jaw after orthodontictreatment, 3 years after concluding RME. Previousauthors reported a slight maxillary displacement

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 109, No. 4

forward (never greater than 1.5 mm, according toWertz"); nevertheless, in some cases the maxilla isseen to recede. According to Carrefio and Menen-dez'? and Bishara and Staley,? the final position ofthe upper jaw is unpredictable. Wertz" in his thirdrecording, 3 months after retention, described areturn of point A to its initial position in 50% ofthe cases. As noted, no changes in the anteropos-terior position of the maxilla were noted in oursample 3 years after treatment. This suggests that,if changes did happen just after RME, they disap-peared later on as the maxilla continued to returnto its initial position after rapid expansion.

In our sample, facial depth underwent statisti-cally significant changes on average. The 10 in-crease observed implies that the mandible ad-vanced with respect to nasion, as occurs at theyearly rate of 0.330 in normal growth. Thus, after 3years, the increase with growth equals one fulldegree, as seen in the present study, and cannot beattributed to the RME performed at the beginningof full orthodontic treatment. Most authors agreethat mandibular backward displacement is the re-sult of the posterorotation produced by rapid ex-pansion, with a retrusion of poi~t B. Only Wertz'described a forward displacement of point B in anumber of cases. He claimed that the mandible canbe propelled or retruded with RME, exhibitingmoreover a routine return to the initial positionduring retention, 3 months after concluding rapidexpansion. Our results do not contradict theseobservations.

Facial convexity significantly decreased by 1 mm,on average, after the 3 years of treatment. In thissense, growth also causes a 0.33 mm yearly decrease.This phenomenon is due to the likewise physiologicadvance of the mandible, which inclines the nasion-pogonion plane (facial plane) forward, therebyshortening the distance between the latter and pointA. Several authors-":":" described an advance ofpoint A with RME, measured when concluding su-ture opening. Thus our results regarding this vari-able differ from those of earlier studies. However,this difference could be explained by the fact thatour measurements were preformed at differenttimes than those of earlier studies, i.e., 3 years afteractive growth and not just after RME.

CONCLUSIONS

1. Three years after initiating orthodontic treat-ment, we observed no vertical or anteroposte-rior facial skeletal differences between theinitial and the final records that could be at-

•.

Veldzquez; Benito, and Bravo 367

tributed to the RME, with which orthodontictreatment was initiated.

2. The statistically significant changes seen 3years after taking the initial records wereonly those expected from normal growth.

3. Thus the classically described skeletal alter-ations resulting from RME, such as anterioropen bite and mandibular posterorotation,seem to be compensated for or corrected inthe course of orthodontic treatment. Never-theless, this compensation does not seem tobe a major consequence or effect of treat-ment itself but of function, which allows thatgrowth evolves normally, without great varia-tions. The continued changes would likely bea consequence of normal growth.

4. Future studies made of RME cases, wellafter removal of all retaining appliances,would be of value to determine ultimatestatus of dentofacial components.

REFERENCES

1. Davis W, Kronman J. Anatomical changes induced by split-ting of the midpalatal suture. Angle Othod 1969;39:126-32.

2. Haas AJ. Palatal expansion: just the beginning of dentofa-cial orthopedics. AM J ORTHOD 1970;57:219-55.

3. Wertz RA. Skeletal and dental changes accompanying rapidmidpalatal suture opening. AM J ORTHOD 1970;58:41-65.

4. Bejarano E. Variaciones de la posicion maxilar ocasionadaspor la expansion de la sutura palatina media. Rev EspOrtod 1977;7:151-63.

5. Costa A. Disyuncion maxilar. Rev Esp Ortod 1978;8:247-6l.6. Ferre F. Cambios ortopedicos del maxilar superior en sen-

tido vertical. Ortod Esp 1981;11:25-9.7. Timms DJ. Rapid maxillary expansion. Chicago: Quintes-

sence 1981:91-4.8. Benito E. Un estudio sobre la expansion rapida del maxilar

superior. Ortod Esp 1982;26:41-50.9. Bishara SE, Staley RN. Maxillary expansion: clinical impli-

cations. AM J ORTHOD DENTOFACORTHOP 1987;91:3-14.10. Sarver DM, Johnston MW. Skeletal changes in vertical and

anterior displacement of maxilla with bonded RME appli-ances. AM J ORTHOD DE TOFACORTHOP 1989;95:462-6.

11. Hernandez AM. Disyuncion y anclaje extraoral. Experienciapersonal. Rev Esp Ortod 1990;20:59-77.

12. Carreno J, Menendez M. Cambios maxilares producidos porla disyuncion. Ortod Esp 1991;32:171-82.

13. Silva Filho OG da, Vilas Boas MC, Capelozza Filho L.RME in primary and mixed dentitions. AM J ORTHODDE TOFACORTHOP 1991;100:171-9.

14. Ricketts RM, Bench RW, Gugino CF, et al. Bioprogressivetherapy. Book 1. Denver: Rocky Mountain Orthodontics,1980:55-9.

15. Dalhberg G. Statistical methods for medical and biologicalstudents. New York: Interscience Publications, 1940.Reprint requests to:Dr. Paloma VelazquezOvdn. Dr. Gadea 27-7° B03003-Alican teSpain