Abstract
Objectives
To bring a neonatal classification system of unilateral cleft lip and palate and to correlate this classification with the distribution of the permanent lateral incisor and maxillary growth.
Design
Retrospective with longitudinal follow-up.
Setting
Tertiary.
Patients
A total of 112 individuals with treated unilateral cleft lip and palate and 30 controls.
Main Outcome Measures
Unilateral cleft lip and palate neonatal casts were classified anatomically in four categories, in which Class 1 corresponds to a maxillary arch with a narrow alveolar cleft; Class 2 corresponds to a balanced form; Class 3 corresponds to a wide cleft and short maxilla; and Class 4 corresponds to a wide cleft and long maxilla. The classification was correlated with the distribution of the permanent lateral incisor. Maxillary growth was evaluated using a cephalometric analysis after the age of 10 years.
Results
Clinical classification of unilateral cleft lip and palate found 10 cases of Class 1 (8.9%), 34 cases of Class 2 (30.4%), 46 cases of Class 3 (41.1%), and 22 cases of Class 4 (19.6%). The permanent lateral incisor was most often present in narrower clefts (Classes 1 and 2); whereas, large clefts (Classes 3 and 4) were relatively more frequently associated with an agenesis of the permanent lateral incisor (P = .019). Maxillary growth impairment was most severe in Class 3, with a mean sella-nasion-A point angle at 71.9° ± 4.6° (P < .001).
Conclusions
Using the cleft width, arch form, and shape of the nasal septum, unilateral cleft lip and palate can be classified into four different classes at birth, which can all give information about permanent lateral incisor agenesis and maxillary growth.
All cases of unilateral cleft lip and palate (UCLP) are not similar and do not have the same prognosis. Midfacial growth is multifactorial, and the numerous factors reported include the initial anatomo-clinical form (Johnson et al., 2000; Liao et al., 2010; Tomita et al., 2010; Chiu et al., 2011; Meazzini et al., 2011), the inherent growth potential of the child, and the treatment protocol and surgical techniques (Wiggman et al., 2012), explaining interindividual variations. The general clinical criteria used to establish these variations are based on the extent of the nasolabial deformity, the width of the cleft, and the cleft palate severity. Unfortunately, these criteria, when used alone, give limited information about the prognosis of maxillary growth and the risk of the surgical treatment on maxilla hypoplasia.
Normal dental development is another important outcome factor that will affect the future orthodontic treatment and dental rehabilitation. Agenesis of the permanent maxillary lateral incisor on the cleft side is frequent in UCLP, occurring in about half of the cases (Delestan et al., 1998; Tsai et al., 1998; Lourenço Ribeiro et al., 2003; Tortora et al., 2008). Teeth are also a marker of the embryological development of the maxilla (Couly and Monteil, 1982), and their agenesis can be viewed as sign of a worse prognosis.
To better understand the prognosis and impacts of treatments on the outcomes of UCLP, the relevant anatomo-clinical inherent factors, identifiable at birth, must be known. Thus, the purpose of our work is to establish a neonatal classification system of UCLP that can initially predict the growth potential of the maxilla. Our hypothesis is that the maxillary arch form at birth can potentially predict the maxillary growth potential and can be indicative of an agenesis of the permanent lateral incisor. Specific aims were to correlate the initial deformities with (1) the sagittal growth of the maxilla and (2) the distribution of the permanent maxillary lateral incisor.
Materials and Methods
This retrospective study reviewed cleft patients with strict longitudinal follow-up, from birth until adolescence, treated at the Lapeyronie Hospital, Montpellier, France. All of the consecutively treated cleft lip and palate (CLP) patients, between 1978 and 2000, were reviewed. This research has followed the principles outlined in the Declaration of Helsinki.
The inclusion criteria were UCLP patients with longitudinal follow-up, born at term, and operated upon by the same surgeon with the same treatment protocol (Montoya et al., 2002). The surgical technique and sequence consisted of the Malek protocol with an early soft palate closure at 3 months of age followed by simultaneous closure of the lip and hard palate at 6 months of age. A maxillary plaster cast taken within 1 week after birth and a panoramic radiograph taken at the age of 6 to 7 years, as well as a panoramic and cephalometric radiographs taken after the age of 10 years, were required for the chart to be complete.
Incomplete files, bilateral CLP, patients lost to follow-up, and syndromic cleft patients were excluded from the study. A series of 30 cephalometric radiographs of patients from the same geographic region and free from any craniofacial pathology were used as controls.
Anatomical Classification of the Maxillary Arch at Birth
A morphological diversity of the maxillary arch exists in the UCLP. Using the clinical team experience and observations of plaster casts, the anatomical shape of the maxillary arch of UCLP patients can be classified into four different categories (Fig. 1).
Class 1 corresponds to a maxillary arch with a very narrow alveolar cleft. The two cleft margins are sometimes in closed contact with a tiny bridge.
Class 2 corresponds to a balanced form in which the shape of the maxillary arch is close to the controls. The cleft is narrow, and the small segment (lateral palatal segment) is not displaced, presenting a harmonious curve, without a sagittal shift compared with the large segment.
Class 3 corresponds to a wide and short form. In this case, the transverse distance of the maxillary arch is definitely more important than its sagittal length when compared with the controls. Furthermore, the nasal septum is significantly deviated anteriorly with a torque effect.
Class 4 corresponds to a wide and long form. The transverse distance is close to that of Class 3, but the sagittal length of the arch is increased when compared with the controls. The septum is rectilinear.

Clinical classification of unilateral cleft lip and palate. The Class 1 arch is represented by 1; Class 2 by 2; Class 3 by 3; and Class 4 by 4.
A cleft surgeon and cleft team dentist carried out this classification, which was validated in the first part of our work. Validation was performed after evaluating the concordance of four different examiners who classified 50 casts chosen randomly from the group of 121 patients. Results from the first part of our work validated this clinical classification with a good intraexaminer and interexaminer concordance analysis (κ > 0.6).
Distribution of the Primary and Permanent Lateral Incisor on the Cleft Side
At the age of 6 to 7 years and during adolescence, an oral clinical examination and a panoramic radiograph were performed. The distribution pattern of the primary and permanent maxillary lateral incisor on the cleft side was classified according to Tsai et al. (1998) in which (1) AB = agenesis; (2) X = localized on the large segment, mesial to the cleft; (3) Y = localized on the small segment, distal to the cleft; and (4) XY = duplicated, on each side of the cleft. The form of the primary or permanent lateral incisor on the cleft side was not studied.
Cephalometric Analysis
After the age of 10 years, the cephalometric radiographs of the UCLP patients and the 30 controls were analyzed. A classical analysis was obtained using Procuste software (Procuste sarl, Caen, France) (Fig. 2). The anteroposterior position of the maxilla was evaluated using the sella-nasion-A point (SNA) angle. Vertical measurements evaluated were the Frankfort-occlusal plane angle (Fr/Op), and the Frankfort-mandibular plane angle (FMA). The sella-nasion-B point (SNB) and A point-nasion-B point (ANB) angles were measured also.

Cephalometric analysis.
Patients who underwent maxillary advancement surgery were noted. The type of maxillary surgery was also categorized into LeFort I; internal distraction; and rigid external distraction (RED).
Statistical Analysis
All variables were divided into continuous and categorical variables. Categorical variables were further subdivided into dichotomous and polychotomous variables. The Pearson chi-square test was used to analyze two dichotomous or polychotomous categorical variables or a dichotomous categorical variable with a polychotomous categorical variable. The t test was used to evaluate a dichotomous categorical variable with a continuous variable. An analysis of variance was chosen to study polychotomous categorical variables with continuous variables. Regressions were used to investigate the effect of a variable while controlling for the influence of other variables. For all statistical tests, a 95% confidence interval (CI) was used, and P < .05 was considered statistically significant.
All statistical analyses were realized using SAS software (Statistical Analysis System, Institut Universitaire de Recherche Clinique, Montpellier, France), the SPAD software (Système Portable pour l'Analyse des Données, Institut Universitaire de Recherche Clinique, Montpellier, France), and the BESPIM software (Biostatistique Epidémiologie Santé Publique et Information Médicale, Nîmes, France).
Results
Anatomical Classification of the Unilateral Cleft Lip and Palate
Of 713 clefts operated upon from 1974 to 2000, there were 399 cases of UCLP, and a total of 112 charts met the inclusion criteria (70 boys, 42 girls).
The clinical classification of UCLP maxillary arches at birth found 10 cases of Class 1 (8.9%), 34 cases of Class 2 (30.4%), 46 cases of Class 3 (41.1%), and 22 cases of Class 4 (19.6%) (Table 1).
Group Comparison
Statistically significant.
The age at time of cephalometric evaluation differed significantly between the groups due to the younger control group (Table 1). No statistically significant differences in age were found between the classes when looking only at the UCLP patients (P = .409).
The sex distribution differed between the groups (P = .011). The control group had a normal distribution (skewness = 0.00). Boys were more frequent in Class 1 (skewness = 0.48), Class 2 (skewness = 0.95), and Class 4 (skewness = 2.28). Girls were slightly more frequent in Class 3 (skewness = -0.18).
Association Between the Arch Class and the Distribution of the Permanent Maxilary Lateral Incisor on the Cleft Side
Results of associations between the anatomical class and the distribution pattern of the permanent lateral incisor (independent to its form) are shown in Table 2. On the cleft side, the permanent lateral incisor was absent in 38% (n = 43), duplicated in 7% (n = 8), localized on the large segment in 7% (n = 8), and localized on the small segment in 47% (n = 53). This distribution differed according to the anatomical class of the maxillary arch. Large clefts (Classes 3 and 4) were more often correlated with an agenesis of the permanent maxillary lateral incisor (agenesis in 47.1%); whereas, narrower clefts (Classes 1 and 2) were commonly associated with the presence of the permanent maxillary lateral incisor (agenesis in 25.1%; P = .019), which was often located on the lateral palatal segment. All patients from the control group had both permanent lateral incisors present.
Association Between Maxillary Arch Class and the Distribution of the Permanent Maxillary Lateral Incisor on the Cleft Side *
X = localized on the large segment, mesial to the cleft; Y = localized on the small segment, distal to the cleft; XY = duplicated, on each side of the cleft; AB = agenesis.
The association between the arch class and the distribution of the primary lateral incisor followed the results reported in the first part of our work (Table 3). In summary, the primary lateral incisor was usually duplicated in Class 1, located on the lateral palatal segment in Classes 2 and 4, and absent in Class 3 (P < .001). When the primary maxillary lateral incisor on the cleft side was absent, the permanent maxillary lateral incisor was usually absent also (85.7%).
Association Between Maxillary Arch Class and the Distribution of the Primary Maxillary Lateral Incisor on the Cleft Side †
X = localized on the large segment, mesial to the cleft; Y = localized on the small segment, distal to the cleft; XY = duplicated, on each side of the cleft; AB = agenesis.
Statistically significant.
Association Between the Arch Class and the Maxillary Growth
The SNA angle differed significantly between the classes (P < .001) (Table 4; Fig. 3). A regression analysis also showed that the neonatal class had a significant effect on the SNA angle, after controlling for age, sex, and the distribution of the primary and permanent lateral incisors (β = -1.58; 95% CI = -2.43 to -0.74; P < .001). Maxillary retrusion was most severe in Class 3. The mean SNA angle was 71.9° ± 4.6°, which was significantly smaller than in the controls (P < .001) and all other UCLP classes (P < .03). The maxillary deficiency was present but less severe in Classes 1 and 4, with significantly smaller SNA angles compared with the controls (P < .01). The Class 2 cases had the best growth potential, with a SNA angle not significantly different than that of the controls (P = .322). The ANB angle followed a similar pattern, with significant differences between the groups (P < .001). The ANB angle was the lowest in Class 3 (-1.2° ± 4.5°), followed by Class 4 (-0.1° ± 3.5°), Class 1 (1.6° ± 2.8°), and Class 2 (2.6° ± 3.0°). The SNB angle also differed significantly between the groups (P < .001), with the lowest mean value in Class 3 (73.0° ± 3.5°), followed by Class 1 (73.9° ± 3.8°), Class 2 (75.8° ± 4.0°), and Class 4 (76.0° ± 3.3°). Within vertical measurements, the Fr/Op angle did not differ significantly between the groups (P = .236). The FMA was borderline significant due to the tendency for a slightly increased lower facial height in the cleft patients compared with the controls (P = .054).
Association Between Maxillary Arch Class and Cephalometric Measurements (Mean ± Standard Deviation)
Statistically significant.
Note: SNA = sella-nasion-A point angle; SNB = sella-nasion-B point angle; ANB = A point-nasion-B point angle; Fr/Op = Frankfort- occlusal plane angle; FMA = Frankfort-mandibular plane angle.

Association between maxillary arch class and the SNA angle.
The agenesis of the permanent lateral incisor was associated with a smaller SNA angle (75.0° ± 5.1° with agenesis; 76.8° ± 5.1° without), but the difference was not significant (P = .091). Similarly, the agenesis of the primary lateral incisor was also associated with a smaller SNA angle (74.5° ± 4.4° with agenesis; 76.6° ± 5.2° without), but the difference was again not significant (P = .109). The FMA and Fr/Op angles were not affected by the agenesis of the primary or permanent lateral incisors.
Requirements for maxillary surgery differed between the groups (P < .001) (Table 5). Maxillary advancement surgery was most often performed in Class 3 cases (28%), usually with the use of RED (54%). The balanced Class 2 had the lowest requirements (3%), followed by Class 1 (10%) and Class 4 (18%), all usually treated with a LeFort I.
Association Between Maxillary Arch Class and Maxillary Advancement Surgery
RED = rigid external distraction.
Statistically significant.
Discussion
Neonatal maxillary arches of UCLP can be classified in four different classes based on the width of the cleft, the arch form (length and width), and the curvature of the nasal septum (Table 6). Class 1, with a narrow cleft, is associated with an intermediate maxillary growth potential, and a permanent maxillary lateral incisor (on the cleft side) most often present. The balanced Class 2 presents most often a normal maxillary growth, the presence of a permanent maxillary lateral incisor, and the lowest requirements for maxillary advancement. The wide and short (with a curved septum) Class 3 is the most severe form, with an impaired maxillary growth, an often absent permanent maxillary lateral incisor, and the highest requirements for maxillary advancement. These patients require a very close follow-up during their mixed dentition to maximize their growth potential. Finally, the wide and long (with a straight septum) Class 4 is associated with an intermediate maxillary growth, and a permanent maxillary lateral incisor most often absent. The clinical interest of this neonatal clinical classification was validated in the first part of our work by good intraexaminer and interexaminer concordance and has been shown to be practical, easy, and reproducible.
Neonatal Class Predictive Tendency
All forms of UCLP are not identical. This probably has an influence on growth potential, regardless of the treatment protocol. Using the width of the cleft as the only prognostic factor is insufficient, and numerous reports in the literature are contradictory. Suzuki et al. (1993) evaluated 39 UCLP patients at 4 years of age and reported no relationship between cleft size and later forward growth of the maxilla. Similarly, other studies (Johnson et al., 2000; Meazzini et al., 2008; Meazzini et al., 2011) showed no correlation between the width of the cleft and maxillary growth at 5 to 6 years of age. Long-term results from Wiggman et al. (2012) also found no association between the initial width of the cleft and the anteroposterior maxillary growth, after evaluating 45 UCLP patients at 17 years of age. Other studies showed the opposite (Liao et al., 2010; Tomita et al., 2010; Chiu et al., 2011). Liao et al. (2010) found a significant correlation between the cleft size at the time of palate repair and the maxillary length and protrusion. Of the 39 UCLP patients studied, those with a large cleft had a shorter and more retrusive maxilla than those with a small cleft by the age of 9 years. In another study of 29 UCLP patients at 9 years of age, Chiu et al. (2011) reported that patients with a small cleft area have a more protruded maxilla than those with a large cleft area.
Using the form (size and shape) of the arch, the width of the cleft, and the curvature of the nasal septum seem to be more relevant. In line with our study, Peltomaki et al. (2001) found a great variation in arch size when studying neonatal casts. They distinguished two types according to the width of the cleft, and the length and circumference of the arch. Growth potential at 5 to 6 years appeared to be compromised in wide clefts associated with a hypoplastic arch (smaller length and circumference), compared with narrow clefts with no maxillary hypoplasia. Our neonatal classification was also previously evaluated by another type of cephalometric study and showed similar results, with Class 3 again showing impaired maxillary growth (Mattei et al., 2002).
Orthodontic stability also appears to be influenced by the maxillary arch form in UCLP cases. Al-Gunaid et al. (2008) identified three forms of arch in UCLP. The balanced symmetrical forms have a good prognosis. A risk of relapse exists at the intercanine level in forms with unilateral collapse. A global risk of relapse is present in forms with bilateral collapse. Reiser et al. (2010) found an association between the cleft size and crossbite in children with UCLP. Large cleft sizes at the level of the cuspid points in infancy were significantly associated with less anterior and posterior crossbite at 5 years.
The distribution pattern of the permanent maxillary lateral incisor on the cleft side found an agenesis in 38%, a duplication in 7%, a lateral localized on the large segment in 7%, and a lateral localized on the small segment in 47%, which agreed with prior reports (Bohn, 1963; Ranta, 1971; Suzuki and Takahama, 1992; Tsai et al., 1998). Tsai et al. studied 137 UCLP patients and found a comparable distribution: The permanent lateral incisor was absent in 51%, duplicated in 1%, localized on the large segment in 2%, and localized on the small segment in 46%.
For teeth at the cleft margins, a positive correlation was previously reported between the width of the cleft and dental agenesis (Ehmann et al., 1976). Our results agreed with this observation; we found that the wide forms (Classes 3 and 4) were more often associated with an agenesis than the narrower forms (Classes 1 and 2) (P = .019). The embryologic relation between these dental agenesis in clefts and skeletal malformations has been recognized for a long time (Tondury, 1950; Pfeifer, 1966; Zilberman, 1973; Ehmann et al., 1976). Our study found that the agenesis of the permanent or primary maxillary lateral incisors on the cleft side had a tendency to be associated with a smaller SNA angle in adolescence, but this association could not be confirmed because it was not statistically significant (P = .091 and .109, respectively). The theory stating that the missing lateral incisor can be considered as the malformative expression of a localized hypoplasia was also validated with prior reports (Meazzini et al., 2008; Meazzini et al., 2011). Meazzini et al. (2011), in a study of 129 UCLP patients at 5 years of age, reported that congenitally missing laterals, a sign of inherent tissue hypoplasia and lack of migration of neural crest cells, seem to be the most important noniatrogenic factor linked with the maxillary growth potential (P < .001).
Our study has several limitations. The design of this study was retrospective and not randomized, introducing a selection bias. The examiners were also not blinded, which could have introduced some measurement bias. The sample size was also small in some groups (especially Class 1). Furthermore, the sex distribution differed between the groups, which could possibly have led to the different outcomes. Therefore, the statements about treatment should be taken with caution.
Conclusions
All individuals with UCLP do not have the same prognosis, despite the surgical technique used. It thus seems necessary to identify, within UCLP cases, the subgroups with the worse growth potential in order to treat them appropriately. The proposed neonatal classification of UCLP permits a characterization of the different forms at birth by associating the width of the cleft with the arch form, the curvature of the septum, and the overall sagittal dimension of the maxilla. The identified neonatal class can potentially predict a growth impairment and agenesis of the permanent maxillary lateral incisor on the cleft side. In our opinion, the most important is the identification of Class 3 due to the frequent agenesis of the lateral and the poor growth. At birth, this information can help the parents to better prepare themselves for eventual outcomes and can be used by the cleft team to potentially adapt their treatment plan.
