Minimal incision palatoplasty with or without muscle reconstruction in patients with isolated cleft palate—a cephalometric study at 5 and 10 years

Summary

Objective

To compare differences in facial growth in patients with isolated clefts of the hard and/or soft palate treated with the minimal incision technique without (MI) or with muscle reconstruction (MMI).

Subjects and method

A consecutive series of 170 Caucasian children born with isolated cleft palate were studied. Individuals with other craniofacial malformations, apart from Pierre Robin sequence (PRS), were excluded. The patients were treated surgically with MI (n = 85) or MMI (n = 85) palatoplasty (mean age: 13 months) and divided further into two subgroups: clefts within the soft palate only (small cleft, n = 51) and within the hard and soft palate (big cleft, n = 119). A retrospective evaluation at 5 (mean 5.4) and 10 (mean 10.3) years was performed using lateral cephalograms. Twelve skeletal and one soft tissue measurement was evaluated. Both 95% and 99% confidence intervals were calculated, two-way ANOVA and mixed model analysis was performed including/excluding PRS.

Results

At 5 years, statistically significant increased inclination of the palatal plane in the big MMI cleft group (P < 0.01), increased posterior upper face height (P < 0.01), and longer mandibular length (P < 0.001) in the small MI cleft group was observed. At 10 years, statistically significant increased inclination of the palatal plane (P < 0.001), decreased posterior upper face height (P < 0.001), and longer palatal length (P < 0.01) was seen in the big MMI group.

Limitations

Retrospective single centre study, limited sample size, three surgeons.

Conclusion

Minor differences in craniofacial morphology were found between patients with isolated clefts treated with MI or MMI technique and between small and big cleft lengths.

Introduction

Historically, children born with an isolated cleft palate (ICP) are generally treated surgically at the age of 6 to 36 months. The discrepancy is corrected in one or two stages by the application of various surgical techniques (1–3). The aim of the operation, independent of the chosen technique, is closure of the cleft which improves function in both oral and nasal cavities as well as speech. Unfortunately, a scar tissue is formed simultaneously. This side-effect has been blamed for restriction in later craniofacial growth, mainly of the maxilla (4).

A one-stage palatoplasty for closure of the cleft in the palate called minimal incision (MI) palatoplasty (Figure 1a) was introduced in the Stockholm Craniofacial Team in 1987 (5) in order to minimize the creation of scar tissue. A similar technique was described by Mendosa in 1994 (6). Sommerlad introduced a detailed reconstruction of the muscles of the soft palate during palatoplasty claiming an improved speech development (7–9).

The MI technique was further developed with the addition of muscle reconstruction (MMI) (Figure 1b) in 1997 in order to improve speech (10). From a previous study of dental casts and medical records where the two techniques were compared, no significant differences in transversal or anteroposterior growth of the dentition was found at 5 years (11). In a study, where speech was investigated, the MMI technique was not found significantly superior to the MI technique but resulted in fewer velopharyngeal flaps (10). A more recent investigation in children treated with MMI, from the same research group, revealed an improvement in hypernasality, weak pressure consonants, deviant s-articulation, and velopharyngeal function, while significant deterioration was found for audible nasal air leakage from 5 to 10 years. A need for additional secondary velopharyngeal surgery after 10 years of age was suggested (12). It is not known whether facial growth in children operated with the MI technique remains the same as in children operated with the MMI technique at the age of 10 years.

According to previous studies on growth of surgically treated children born with an ICP, surgery of the hard palate inhibits forward translation of the maxilla and forward development of the alveolar process, while surgery of the soft palate might inhibit posterior vertical maxillary development (13). Comparing older one-stage surgical techniques such as Wardill-Kilner and von Langenbeck, the effect on anteroposterior and vertical growth of the craniofacial complex has not been found to differ (14, 15). In a material of Japanese adults with ICP, shorter maxillary length and a more posteriorly positioned nasomaxillary complex compared with patients without a cleft was reported (16). In children treated with the MI technique or the Veau-Wardill-Kilner technique at the age of 5 and 10 years, only minor differences in growth of the craniofacial complex have been reported, independent of cleft length or inclusion of children with Pierre Robin sequence (PRS) (17).

The aim of this study was to compare the MI technique with or without muscle reconstruction regarding facial growth at 5 and 10 years of age. The null hypothesis was defined as there is no difference in facial growth in patients born with isolated clefts of the hard and/or soft palate surgically treated with the MI technique with or without muscle reconstruction.

Materials and methods

This study was approved by the ethical committee in Stockholm, nr 2009/1680-31/2.

Patients

Between the years 1987 and 2005, 185 Caucasian, non-syndromic (including PRS) children were born with ICP in the region of Stockholm and treated with one of the two investigated palatoplasties (MI 1987–96; MMI 1997–2005) (Figure 1). Fifteen children were excluded from the present study (Figure 2) due to late surgery (n = 1), few palatoplasties performed by low-volume surgeon (n = 3; one patient treated by a surgeon performing the MI technique and two patients treated by another surgeon with the MMI technique), death before the age of 5 years (n = 4), or missing records (n = 7). Thus, the final material included 170 children.

The patients were divided into two groups depending on the surgical procedure applied for closure of the cleft in the palate (MI: original or MMI: modified MI technique). Each group was further divided into two subgroups according to the sagittal extension of the cleft (1: small or 2: big). One of the main surgeons who treated both surgical groups categorized the clefts as small or big based on the medical records, where cleft length was noted down in more than two categories during conference meetings and operation, and the casts of the patients. This was performed at an earlier study.

Group MI1

Clefts repaired using the MI technique and extending only within the soft palate and a notch less than 3 mm in the posterior border of the hard palate, n = 30; 1 PRS (16 boys, 14 girls).

Group MI2

Clefts repaired using the MI technique and extending within the soft and hard palate, n= 55; 11 PRS (24 boys, 31 girls).

Group MMI1

Clefts repaired using the MMI technique and extending only within the soft palate and a notch less than 3 mm in the posterior border of the hard palate, n = 21; 0 PRS (12 boys, 9 girls).

Group MMI2

Clefts repaired using the MMI technique and extending within the soft and hard palate, n = 64; 9 PRS (29 boys, 35 girls).

Treatment methods

Three experienced surgeons treated groups MI1 and MI2 surgically at the department using the MI technique. The average age was 12.6 (SD = 1.3) months. Two of the surgeons also treated all patients in groups MMI1 and MMI2 with the MMI technique. The average age in the groups at the time of surgery was 12.7 (SD = 1.5) months.

Examination methods

A cephalometric investigation of 294 available lateral cephalograms was performed at 5 (mean age 5.4 ± 0.6 years) and at 10 years (mean age 10.3 ± 0.6 years) of age (Table 1). One hundred twenty-four of the 170 children were examined at both 5 and 10 years of age. Analogue radiographs were scanned at 300 dpi and an 8-bit greyscale with a scanner (Epson Perfection V700 Photo, Seiko Epson Corp., Japan) providing digitized cephalograms. The magnification was adjusted to zero for all radiographs. Cephalometric analysis was performed using cephalometric software (Viewbox 4.0.1.7, dHAL Software, Kifissia, Greece). Computerized tracing technique has been shown to be less time consuming and equally reliable as hand tracing (18). All cephalograms were digitized and analysed at 5 and 10 years by one author (KP). Reference points defined by Solow and Tallgren (19) and Legan and Burstone (20) were identified directly on the digitized radiographs. For the calculation of face heights, one constructed plane (HP: a line −7^o from n−s line passing through s) and three constructed points (sp′: the perpendicular projection of sp on n−gn line, pm′: the perpendicular projection of pm on HP, tgo: the intersection between ar-rli and gn-mlp) were additionally used. Twelve skeletal and one soft tissue measurement was evaluated (Figure 3).

The intraobserver errors of measurements were calculated by selecting 15 random cephalograms from each surgical group and digitizing them twice within a 2-week interval. After this step, all 294 radiographs were unidentified and randomly traced twice by the same author (KP) and the midpoint from the two digitizations for every reference point was used. The quality of the lateral cephalograms varied, and it was difficult to provide full registrations for all patients. Consequently, all available radiographs were included, but measurements based on landmarks hard to identify were excluded for every individual film. The main area of uncertainty in some radiographs was the area of the anterior nasal spine (in particular the ss point) and the soft tissue profile.

Statistical methods

The data at 5 and 10 years were analysed separately to compare the two techniques and the two cleft lengths by using two-way analysis of variance (ANOVA) with the factors surgery types (MI, MMI) and cleft types (1, 2). To reveal growth differences from 5 to 10 years, a mixed model analysis was performed with the additional factor of time (5 or 10 years) (21, 22). 95% confidence intervals were calculated. Due to multiple testing, the level of significance was set to 0.01 and confidence intervals were calculated at 99%. The statistical analyses were repeated excluding patients with PRS, as well as investigating only the 124 children providing cephalograms at both 5 and 10 years.

The statistics were calculated using the software SPSS 23.0 (IBM Corp., New York, USA).

The intraobserver method errors S_i were calculated using the formula $Si=±∑d22n$⁠, where d is the difference between the first and second measurement and n is the number of double registrations (23). The method error varied from 0.38 per cent for n−sp′/n−gn to 1.55^◦ for ML/RL (Table 2).

Results

Complete data

At 5 years (Table 3), a statistically significant longer mandibular length (P < 0.001) and increased posterior upper face height (P < 0.01) in the MI small cleft group compared with the MI big cleft group and an increased inclination of the palatal plane (NSL/NL) in the MMI big cleft group compared with the MI big cleft group (P < 0.01) was seen. An increased inclination of the palatal plane (NSL/NL) in the MMI big cleft group compared with the MMI small cleft group, as well as an increased posterior upper face height in the MI small cleft group compared with the MMI small cleft group was also found (P < 0.05). The longer mandibular length (P < 0.01) and the increased posterior upper face height (P < 0.05) in the MI small cleft group compared with the MI big cleft group were less significant when excluding children with PRS.

At 10 years (Table 4), a statistically significant increased inclination of the palatal plane (NSL/NL), a decreased posterior upper face height, and a longer palatal length in the MMI big cleft group compared with the MI big cleft group (P < 0.001, P < 0.001 and P < 0.01 respectively) was seen. In addition, an increased inclination of the palatal plane (NSL/NL) in the MMI small cleft group compared with the MI small cleft group, an increased upper face height ratio (n−sp′/n−gn) in the MMI group compared with the MI group for both small and big subgroups and an increased posterior upper face height in the MMI small cleft group compared with the MMI big cleft group was observed (P < 0.05). The difference in the upper to total anterior face height ratio (n−sp′/n−gn), where MI big to the MMI big subgroups was compared, showed no significance when excluding patients with PRS.

From 5 to 10 years (Table 5), the main significant finding was the growth changes with time. A significant increase was observed at variables NSL/NL, NAPg, palatal plane length, mandibular length, n−sp′/n−gn, and both posterior face heights, and a significant decrease at the variables SNA, ANB, NSL/ML, and the gonial angle (P < 0.001). Concerning type of surgery, statistically significant differences in results were found showing a higher increase in the NSL/NL angle (P < 0.001) and in the n−sp′/n−gn ratio (P < 0.01) in the MMI group. The same results were noted when patients with PRS were excluded.

Longitudinal data

Analysing only the 124 pairs of radiographs, identical results with those of the complete data were provided with the following exceptions: at 5 years, the difference in the mandibular length (P = 0.003) and the posterior upper FH (P = 0.027) for the MI technique between small and big cleft groups decreased in significance. At 10 years, the difference in the palatal plane length was not statistically significant anymore for the big clefts between the MI and the MMI technique.

Discussion

The present study is a retrospective evaluation of growth as measured on digitized radiographs from children born with an isolated cleft in the palate. The children were treated surgically with a palatoplasty at the age of 13 months. Lateral cephalograms were taken at the ages of 5 and 10 years. The surgical technique involving muscle reconstruction implies an extended surgical procedure (11) and consequently increased staff and facilities. Training in performing the additional muscle reconstruction was also needed for the surgeons. The main question was whether the muscle reconstruction benefits the patient enough to motivate the extended surgical procedure. Previous studies comparing the MI and the MMI technique evaluating speech, dentoalveolar changes on dental casts, and surgical variables from medical records reported no significant differences, apart from decreased need (less than 50 per cent) for further surgical correction of velopharyngeal insufficiency when applying the MMI technique (10, 11).

The recorded cephalometric measurements for the total material are in accordance (deviating up to 1.2 degree at mean value) with those on patients with an isolated cleft in the palate at 5–6 and 9–11 years treated with a two-stage technique (24). The main difference is the smaller NSL/ML angle in this material by 2.9 at 5 years and 2.0 degrees at 10 years as a mean. Compared with other studied materials on isolated clefts in the palate at 6 years of age treated with a one-stage procedure (25, 26), the cephalometric values that differ are the smaller NSBa angle (3.2 and 3.5 degrees in the mean value), the bigger SNA angle (1.9 and 2.5 degrees), the smaller NSL/ML angle (2.4 and 3.3 degrees), the smaller length of the palatal plane (1.5 and 2.1 mm in mean value) and of the mandible (3.2 and 4 mm) in our material. The above studies do not report whether the lateral cephalograms were corrected for magnification or not, whereby, the palatal and mandibular lengths measured in our material could be similar.

It is unclear whether the significant differences observed in groups with small and big clefts at 5 years resulted from the differences in length of the cleft at birth or from the extended surgical procedure in the big cleft group. Palatoplasty of longer clefts in the palate could restrict the vertical growth of the palate posteriorly, as found by the decreased posterior upper face height in the big cleft group but only in the MI surgical group. This may result in an increased inclination of the palatal plane to the anterior cranial base. If longer clefts in the palate are related to shorter mandibles, as a shorter mandible may have caused a retropositioning of the tongue resulting in a larger cleft, or the shorter mandible is the result of the surgical treatment cannot be derived from this study. The higher number of children with PRS in the big cleft group did not influence the difference found in the mandibular length as shown from the statistical analysis by including and excluding these patients. The finding regarding length of the cleft in a Japanese material at 9 years of age, that the longer the length of the cleft in the palate before the palatoplasty is, the shorter the anteroposterior maxillary length and the more posteriorly positioned the maxilla will be (27), was not confirmed.

When the two surgical techniques were compared, statistically significant differences were found at the age of 10 years: the muscle reconstruction seems to lead to a restriction of the vertical growth of the palate posteriorly and a relative increase of the anterior mid-face height. Comparing treated Japanese patients with submucous ICP to untreated individuals with submucous ICP and to a non-cleft group of patients, an increased inclination (clockwise rotation) of the palatal plane at 9 and 14 years, a smaller posterior upper face height, and similar anterior upper face height were shown in the treated group (28). Consequently, the surgical intervention in the soft palate could explain the restriction of the posterior area of the maxilla vertically. This is in accordance with our finding of an increased NSL/NL angle in the MMI group. When comparing only the big cleft groups, a decreased posterior upper facial height in the MMI group was seen. Even if a difference regarding vertical growth of the posterior palate was found (palatal plane inclination, posterior upper facial height), no sagittal maxillary growth difference between the two surgical groups was noticed anteriorly as the SNA values were similar. Interestingly, a longer palatal length was found when comparing the big cleft groups at the age of 10 years. This difference was close to the limit of significance at 1 per cent and was not present at the age of 5 years. The increased inclination of the Nasal Line in combination, in the big cleft group, with a longer maxillary length seems to lead to a similar sagittal position of the maxilla anteriorly. Further studies in adolescence and a separation of groups according to gender might clarify whether a change in sagittal dimension is present. Another explanation for the increased NSL/NL angle in the MMI group, combined with the result of a relative increase of the anterior mid-face height, could be the improved vertical growth of the maxilla anteriorly or increased horizontal growth of the mandible.

Historically, data from earlier studies (16, 29, 30) on the effect of palatoplasties in the treatment of patients with an ICP showed a restricted growth mostly seen in the maxilla in the anterioposterior direction. In the present study, no difference in the anterioposterior growth of the maxilla was found. This is in accordance with a previous cephalometric study, comparing the MI to the Veau-Wardill-Kilner technique at 5 and 10 years, where no differences in maxillary growth were observed (17).

When comparing our material on facial growth between the ages of 5 and 10 years with individuals treated with other palatoplasties or born without a cleft, statistically significant changes follow the pattern of growth seen in a previously mentioned study of a two-stage technique (24) and that of normal growth (31).

This retrospective material was collected from a single centre providing care to all children born with a cleft in the Stockholm region. Treatment was provided by a specialized team following the same protocol for the two investigated periods. Three experienced surgeons performed the surgical treatments. All used the same methods.

Ideally, the effect of palatoplasty should be evaluated on material of patients treated by the same operator as the surgeon is an important factor (13). The available material was collected during a number of years due to the limited number of patients, as is the case at most craniofacial centres. The cephalograms were taken according to the national treatment protocol of patients with ICP, and during the collection of the material, not all cephalograms were retrievable or found for every child. Consequently, small subgroups were formed, mainly for the small clefts and especially for the clefts in the soft palate only operated with the addition of muscle reconstruction. Therefore, results for the last group should be interpreted with caution.

The intraexaminer error in this paper was found smaller than that of a previous paper (17) which could be attributed to the better quality of the lateral cephalograms as more recent radiographs were used. Orthodontic intervention in the investigated period between 5 and 10 years may have influenced the values for the measurements at 10 years. Another source of bias may be the different postoperative appearance of fistulas (6 per cent for the big cleft group versus 2 per cent for the small cleft group, 6 per cent for the MI group versus 4 per cent for the MMI group) or the different need for velopharyngeal flap surgery (25 per cent for the big cleft group versus 4 per cent for the small cleft group, 28 per cent for the MI group versus 13 per cent for the MMI group) among groups which might compromise growth (11).

In a previous study, two statistical analyses (three-way ANOVA and mixed model analysis) were used for changes in values between ages 5 and 10 years. These gave equal results (17). Therefore, in the present study, only the mixed model analysis was applied. Additionally, statistical evaluation was performed including and excluding children with PRS. This showed no changes in significant differences, apart from an increased upper face height ratio in the MMI group compared with the MI group for big clefts at 10 years. Similar results not influenced by the inclusion or exclusion of children with PRS were also found in a previous study where the MI technique was compared with the Veau-Wardill-Kilner technique (17).

The main reason for performing the reconstruction of the muscles of the soft palate has been the effect of the surgical procedure on speech. The results in the present study, in combination with those in previous investigations on dental casts, medical records, and speech outcome (10, 11), do not provide enough evidence for superiority of the modified technique (MMI), except for a decreased need for velopharyngeal surgery. Future prospective investigations on larger samples, also including cease of growth, isolation of influencing factors, and comparison with a control material are needed for confirmation and validation of the results from this paper.

Conclusion

There is a minor difference in craniofacial morphology between patients treated with the original MI and the modified MI with muscle reconstruction technique and between small and big cleft lengths in patients born with isolated cleft in the palate. At 5 years of age, an increase in the inclination of the palatal plane to the anterior cranial base, a decreased posterior upper face height, and a shorter mandibular length were found in the big cleft group. At 10 years of age, an increased inclination of the palatal plane, a decreased posterior upper face height, and a longer palatal length were found in the MMI group. The null hypothesis was therefore rejected.

Conflict of interest

None declared.

Acknowledgements

We would like to thank orthodontist Margareta Larson, the Eastman Institute, for her useful ideas and suggestions, statistician Elisabeth Berg, Karolinska Institutet, for her precious help and guidance, and Liisi Raud Westberg, Stockholm Craniofacial Team, for the illustrations of Figure 1.

References