Abstract
Objective
To evaluate lip scarring and the three-dimensional (3D) lip morphology following primary reconstruction in children with unilateral cleft lip and palate (UCLP) relative to contemporaneous noncleft data.
Design
Retrospective, cross-sectional, controlled study.
Setting
Glasgow Dental Hospital and School, University of Glasgow, U.K.
Patients and participants
Three groups of 10-year-old children: 51 with UCLP, 43 UCL (unilateral cleft lip), and 68 controls.
Methods
Three-dimensional images of the face were recorded using stereo cameras on a two-pod capture station, and 3D coordinates of anthropometric landmarks were extracted from the facial images. A novel method was applied to quantify residual scarring and the associated lip dysmorphologies. The relationships among outcome measures were investigated.
Results
Residual lip dysmorphologies were more pronounced in UCLP cases. The width of the Cupid's bow was increased due to lateral displacement of the Christa philteri left (cphL) in both UCL and UCLP patients. In the upper part of the lip, the nostril base was significantly wider in UCLP cases when compared with UCL cases and controls. Scar redness was more pronounced in UCL than in UCLP cases. No relationship could be identified between lip scarring and other measurements of lip dysmorphology.
Conclusions
Stereophotogrammetry, together with associated image analysis, allow early detection of residual dysmorphology following cleft repair.
In the field of cleft-related facial morphology, there has been a lack of controlled studies in younger children to assess the outcome following surgical repair. This is due to the fact that children do not cooperate easily with direct anthropometry, thus preventing the collection of large series of normative data. The innovation of indirect anthropometry provided an ideal opportunity to capture the face in three dimensions (3D) by overcoming the problem of patients' cooperation associated with direct anthropometry. Laser scanning has been utilized for the 3D analysis of facial morphology in patients with cleft lip and palate and control subjects aged 8 to 11 years (Duffy et al., 2000).
The most obvious residual dysmorphology following the surgical repair of unilateral cleft lip and palate (UCLP) is usually seen around the nasolabial region. Several methods have been used to assess the surgical outcome following cleft repair. These have generally been subjective (McDade et al., 1991; Morrant and Shaw, 1996; Al-Omari et al., 2003, 2005) and of limited clinical application due to lack of reproducibility.
Several methods of analysis have been utilized in an attempt to quantify dentofacial dysmorphology in 3D. A computer-based method was developed to measure facial asymmetry from standardized photographs of the nasolabial area following surgical repair of UCLP (Coghlan et al., 1987, 1993). Photographs were taken at a certain angle; this was limited to two dimensions and therefore did not provide a full description of lip morphology. The chosen views also gave limited information about the shape and size of the nasolabial complex.
Minimizing upper-lip scarring following cleft repair is an important outcome measure. Scarring was recently assessed in 20 patients who had unilateral clefts repaired using the Millard technique (Christofides et al., 2006). Patients were evaluated subjectively by means of a questionnaire. Using clinical measurements and photographs, patients were also evaluated objectively by two clinicians. The color, width, texture, and thickness of the scars were assessed using 2D photographs. Measurement details and the reproducibility of the method were not discussed. It was not clear how the color of the scar was measured, and the method of quantifying the scar color difference from the surrounding skin was not described. The study concluded that all the patients had scar transgression of the normal anatomic boundaries of the upper part of the lip.
Aim of the Study
This study was carried out to measure lip scarring and the associated dysmorphology following surgical repair of unilateral cleft lip (UCL) and UCLP cases.
Materials and Methods
Ethical Approval
Approval was granted by the West Glasgow Ethics Committee (REC No. 05/S0709/81). Patients and their parents received an information sheet and signed a consent form. This investigation was conducted on two groups of 10-year-old children: 51 with UCLP and 43 with UCL. All cases have been treated following the same surgical protocol that has been adopted by the managed clinical network in Scotland, CLEFTSiS. At the age of 3 months, a modified Millard cheiloplasty and a McComb primary rhinoplasty were performed. At the age of 9 months, palatoplasty with minimal bony denudation was performed. In addition, 68 noncleft children living in the same geographic area were recruited as a control group.
Data Collection
For each case, the face was captured using the Di3D stereophotogrammetry system (Dimensional Imaging Limited, Glasgow, U.K.). Three-dimensional facial images were taken of each child in a rest position and with maximal smile. Only rest images are analyzed here. The output was in the form of an on-screen colored photorealistic image of the participant in virtual reality modelling language. For right-sided clefts, the 3D images of the face were reflected to the left side to create a homogenous group for statistical analysis. A custom-written computer program allowed us to build the 3D facial models and place facial landmarks (Table 1; Fig. 1). In addition, the following set of linear measurements was used to describe upper lip morphology: sn-ls, cphR-ls, cphL-ls, cphR-sn, cphL-sn, acR-acL, sbalR-sn, sbalL-sn, cR-sn, and cL-sn (Table 2).
The Anatomic Landmarks Digitized on the 3D Model and Used for the Analysis of Upper Lip Morphology in Cleft and Control Cases
Illustration of anatomic landmarks.
Mean and SD of Landmark Pair Distance (mm) †
Significant differences between the mean values.
UCL = unilateral cleft lip; UCLP = unilateral cleft lip and palate.
The errors of the method were assessed by repeat digitization of 11 3D models (five UCLPs and six UCLs). The majority of digitized landmarks were reproducible within a mean of 0.5 mm.
Three methods were applied to assess the residual dysmorphology following cleft repair: lip scarring, examination of a set of linear distances, and principal component analysis.
Lip Scarring
The magnitude of facial scarring was measured using a custom-written program to detect differences in color content between the area of scarring and the surrounding region. For this preliminary analysis, only a single 2D image was used of each participant, taken from the camera with the most orthogonal view of the scar. The cameras were calibrated using standard techniques (Westland and Ripamonti, 2004) to allow the extraction of CIELAB color coordinates from the images. The calibration was based on an image of a Macbeth Color-Checker (GretagMacbeth, Windsor, NY) captured close in time to the corresponding facial capture, so that any day-to-day variations in camera properties and lighting were accounted for. CIELAB color coordinates decompose each color image into three separate images: a luminosity (L*) image, an image based approximately on the relative redness (as opposed to greenness) (a*) and an image based approximately on the amount of yellowness as opposed to blueness (b*) (Fig. 2a). A difference (ΔE*) of 1 between two sets of CIELAB coordinates should correspond to a color difference that is just perceivable (Westland and Ripamonti, 2004). Each transformed facial image was then further analyzed in the following way. First, a region of interest (ROI) of the image was extracted (by hand), which included the base of the nose and the vermilion regions of the lips. This image was then segmented automatically, using the color information, to remove the vermilion section of the lip from the image, so that the ROI was restricted to the region between the base of the nose and the top of the vermilion border (Fig. 2b). The entropy (an image measure related to the variability of the image across space) of the a* image within the ROI was then determined, and the closed region with highest entropy was identified. This was taken to be the candidate “scar” region (Fig. 2c). The color coordinates of this scar region were then compared with those of the surrounding skin.
Objective scar measure.
Linear Distances
A set of linear measurements (Table 2) was applied to measure lip dimensions in UCL and UCLP patients in relation to controls.
Principal Component Analysis
In order to quantify shape, configurations of landmarks were registered across all patients to a common set of axes using generalized Procrustes analysis. This method of superimposition involved three steps: translation, rotation, and scaling. First, the geometric centers of the configurations were found. Then, they were translated so that they were superimposed on their geometric center. Rotation and scaling followed the translation, moving one configuration about its geometric center until the best fit was found between all homologous landmarks preserving the shape information of each configuration.
Principal component analysis (PCA) was applied to the registered data to describe the variation in facial shape by a series of ordered components that capture the amounts of variation from largest to smallest. PCA transforms the data mathematically to a new coordinate system such that the greatest variance by any projection of the data comes to lie on the first coordinate (called the first principal component), the second greatest variance on the second coordinate, and so on. The first principal component corresponds to a line that passes through the mean and minimizes the sum squared error with those points. The second principal component corresponds to the same concept after all correlation with the first principal component has been subtracted from the points. The method provides nonlinear description of morphologic differences between surgically managed cleft cases and the control group.
Results
Lip Scarring
Luminosity was slightly elevated in the scar tissue of UCL cases (Fig. 3). Interestingly, there was little indication of elevated luminosity in the scar tissue of UCLP cases. There is an indication that redness, as measured on the a* color scale, is elevated in the scar tissue of both UCL and UCLP cases. The ratio of the a* value in scar and nonscar areas confirmed the elevated redness of scar tissue and provided a basis for a scar redness index.
Scar measurements (luminosity [L*], a*, and b*).
Linear Distances
In the upper part of the lip, the nostril base was significantly wider on the cleft side (sbalL-sn) in UCLP cases compared with both UCL cases and controls (Table 2; Fig. 4a and b). UCL cases and controls did not significantly differ in terms of nostril base width. There were also significant differences between all three groups in terms of the nasal width (acR-acL). There was an increase in the philtrum width (cphL-ls) on the cleft side, at the crest of the Cupid's bow for both cleft groups. The disparity of the length of the right columella (cR-sn) in UCLP cases was significant in comparison to UCL cases and controls. Both UCLP cases and controls had a mean left columella length (cL-sn) that was significantly longer than in UCL cases.
a: Landmark distances, b: Diagram showing the linear measurements of the nasolabial complex.
There was no significant relationship between scarring measurements and Cupid's bow width for UCL or UCLP cleft cases. Figure 5a and b illustrates the increase in the luminosity of the lip scar for both UCLP and UCL cases. However, no direct relationship exists between the a* score and the width of the Cupid's bow (Fig. 6).
a: Illustration of various shape and luminosity of scarring in UCLP cases, b: Illustration of various shape and luminosity of scarring in UCL cases.
Relationship between the lip scarring and the width of Cupid's bow.
Principal Component Analysis
PCA showed clear differences in lip shape between UCLP cases and control; this was not the case for UCL cases, which showed more overlap with the control group (Fig. 7).
First versus second PC scores by subject group. UCL and UCLP cases and controls are distinguished by circles, triangles, and crosses.
Discussion
The method we used in this study allowed a noninvasive capture of the face and facilitated the 3D objective assessment of the face following cleft repair in 10-year-old Scottish children. Computerized stereophotogrammetry provides a reliable and useful tool for the researcher in the field of cleft-related dysmorphology (Ayoub et al., 1998, 2001, 2003). This imaging modality can be readily applied to the unaffected children, as in this study, which poses no risk (such as radiation) to the subject and costs less time and money than direct anthropometry (Hood et al., 2004; White et al., 2004). The lip measurements provide information about the magnitude and location of residual dysmorphologies following cleft repair.
One of the major advances in understanding cleft lip dysmorphologies is the recognition of the importance of the reconstruction of the orbicularis oris muscle during surgical repair. The orbicularis oris muscle consists of superficial and deep fibers. The deep sphincter part of the muscle does not reach the edges of the interrupted vermilion border. The superficial fibers are misdirected by the cleft. On the lateral side the fibers run vertical to the pyriform aperture. On the medial side they are inserted abnormally, lateral to the anterior nasal spine and floor of the nostril and inserted into the dermis of the cleft edge (Nicolau, 1983). The musculature on the medial side of the cleft is usually underdeveloped and does not extend as far forward to the edge of the cleft as the lateral side.
In Reddy et al. (2008), the postoperative outcomes using the Millard and Pfeifer techniques were compared; linear measurements were obtained both directly on the patients and by using 2D photographs. They found that the Millard incision (based on a rotation flap from the medial side (noncleft side) coupled with an advancement flap on the lateral (cleft side) gave significantly better approximation of the vermilion border, whereas cases that were operated on with a Pfeifer incision resulted in a better length of the lip.
In our study all cleft lip cases were repaired using the Millard method. One of the advantages of this technique is that it allows adjustment as the operation proceeds, with further rotation and advancement movement as required for individual cases. However, incomplete mobilization of the muscle during surgical repair to cross freely to the corresponding end on the opposite side of the cleft could affect lip symmetry. On the other hand, our findings did not agree with shortness of the lip secondary to Millard's repair reported in a number of studies (Fernandes and Hudson, 1993). In rotation advancement repair, most of the tension is in the lower part of the lip. This may have contributed to the increase of the width of the Cupid's bow (cphL-ls), which was detected in both cleft groups in this study. In the upper part of the lip the nostril base was wider on the cleft side in UCLP cases only. This may be secondary to displacement of the alar cartilage, laterally. However, the impact of the deformed bony foundation in UCLP cases may also contribute to the increase of the nostril width on the affected side. This can be confirmed only by radiographic assessment of the maxillary bone height and width, which was not part of this study. The combined effect of the two factors cannot be excluded.
Tamada and Nakajima (2010) examined skin brightness and texture in 26 patients who had Millard's surgical repair of a cleft lip. Each upper lip scar was divided into three portions: upper, middle, and lower, and was evaluated by plastic surgeons who scored each case. The scores were related to the objective measurements. A significant positive correlation was present between the evaluation score and skin brightness in the upper and middle thirds of the lip. These findings are in agreement with the known concept that the scars are more conspicuous on darker skin. The authors related the quality of scarring to the appropriate reconstruction of the orbicularis oris muscle. No information was provided regarding the reliability of the objective method of measuring skin brightness using the facial measurement instrument.
Analysis of scar color using our method is a complex process. The scars tend to be very similar in color to the surrounding tissue, and analysis is complicated by the presence of shadows (which, as they involve mutual illumination of skin are reddish), freckles, highlights (due to moisture on the lips or under the nose), and spots on the skin. We, therefore, developed a technique that searches for image “entropy” in color space. Basically, this can be thought of as chromatic variability. There are still issues with noise from shadows and difficulties with idiosyncratic cases, but we foresee these being dealt with in further refinements to the algorithm. This preliminary statistical analysis suggests that the scar tissue tends to be redder than the nonscar tissue but does not tend to differ in either yellowness or blueness; it also suggests that the size of any difference in luminosity is small, although significant above a ratio of 1 in UCL, but not in UCLP. The scar tissue that was measured in this study extended from the nasal floor through the philtrum down to the vermilion border of the upper lip. This was increased in both UCL and UCLP cases. However, no relationship was detected between the size of Cupid's bow and scar measurements. These findings are different from those reported by Reddy et al. (2008) showing that the scar appeared to be worse when the vertical height of the lip was above the median. This is a logical argument; the scar created would be longer with increased lip length and thus more noticeable. However, our cleft cases did not show statistical differences in lip length compared with noncleft cases.
The possible explanation of the residual dysmorphology in our study is the incomplete approximation of the orbicularis oris muscle during surgical repair. Incomplete advancement of the muscles from the lateral to the medial sides of the cleft results may be the cause in widening of the Cupid's bow which has been seen in our cases. However, this assumption cannot be fully substantiated without magnetic resonance image (MRI) scanning to show muscle dehiscence. Lack of muscle approximation would lead to healing by secondary intention and scar formation. Stretching of scar tissue would lead to widening of the Cupid's bow, which was noted in our cases. Primary repositioning of the orbicularis oris musculature in which the fibers join across the cleft to normal alignment of the lip muscle fibers is essential. The identification of the lump of the levator labii superior muscles and its reconstruction independent from the orbicularis oris muscle may reduce residual lip dysmorphology and improve scarring (Tamada and Nakajima, 2010).
The findings of this study require careful interpretation. Although the measurements are objective, they do not measure the fine esthetic attributes that may also influence the clinical decisions of lip revision. However, there is a general consensus that the objective outcome measures overcome the problems associated with personal opinion and individual perception of facial beauty (Coghlan et al., 1987). It will also overcome poor inter- and intraobserver reliability of the subjective assessment of lip form (McDade et al., 1991; Morrant and Shaw, 1996; Ritter et al., 2002).
This work describes a powerful and objective technique to measure the outcome of the surgical management of cleft dysmorphologies. Objective measures are considered the gold standard, whereby outcomes can be measured and compared beyond personal opinion. Significant residual dysmorphologies were detected following surgical repair of cleft lip and palate; this is due to the combined effect of both the surgical technique and possible alteration in orofacial growth secondary to the cleft.
The benefit of this work will be to deliver, for the first time, a powerful objective method to characterize 3D facial morphology in noncleft and surgically managed cleft children at the age of 10 years. This has the potential to be utilized as a sensitive tool for clinical audit and to measure the surgical outcomes of other surgical procedures. This study is a step in the right direction to improve surgical techniques and to inform clinical decision-making on the possible need of scar revision. Thus, further studies would be carried out to supplement the developed tools with a sound subjective evaluation to develop clear clinical criteria to evaluate patients for surgical revision of lip scarring. Revision of lip scarring is at the discretion of the surgeon and also depends on patients' satisfaction with facial appearance. However, surgeons tend to evaluate the scar more severely than lay assessors and tend to differ in their opinion (Gussy and Kilpatrick, 2006). On the other hand, patients' perception of the severity of the scar is dependent on their previous experience with surgical procedures. Parents of the patients tend to evaluate the severity of the scar more severely than the children. There is a possible lack of consensus among the three parties—surgeons, patients, and parents—regarding the need of revision surgery, and the development of an objective reliable tool would facilitate the decision-making process. Moreover, the results of the study provide information that will help patients and parents have realistic expectations following cleft repair.
Conclusions
This study identified and quantified lip scarring and areas of residual dysmorphology following surgical repair of UCL and UCLP cases. It is suggested that the incomplete approximation of the muscle fibers during surgical repair contribute to the residual deformities of the nasolabial complex.
Footnotes
Acknowledgments
The authors acknowledge the funding from the Chief Scientist Office of Scotland to support this study. We would like to thank the managed clinical network (CLEFTSiS) and the cleft team in Edinburgh and Glasgow for their help in the project. It would not have been possible to conclude this project without the cooperation of the children who took part in this study and the support of their parents.