We calculated the area of the bilateral 30- to 60-degree wedges on the supraorbital measurement plane (Fig. 3). The bandeau ratio for unicoronal patients was the area of the synostotic wedge divided by the nonsynostotic wedge. These wedges encompassed the greatest change in brow shape following surgery. A lower bandeau ratio indicated more severe anterior plagiocephaly. The bandeau ratio in controls was calculated as the left wedge area/right wedge area, with symmetry being a bandeau ratio of 1.
Preoperative asymmetry (time 0) was categorized as mild (bandeau ratio > 0.85), moderate (bandeau ratio of 0.7 to 0.85), and severe (bandeau ratio < 0.7). Immediate postoperative corrections (time 1) were categorized as undercorrected (bandeau ratio < 1.0), overcorrected (bandeau ratio of 1.0 to 1.2), and substantially overcorrected (bandeau ratio > 1.2). Two-year postsurgery (time 2) ratios were categorized as asymmetric (bandeau ratio < 1.0) or symmetric/overcorrected (bandeau ratio >1.0)
Skull Base and Midface Twists
The midface twist angle was measured in the coronal plane between the nasion-prosthion and a line perpendicular to the orbitale-orbitale (Fig. 4). The skull base twist angle was measured on the axial plane as the angle between the sella-opisthion and sella-nasion lines (Fig. 4).
The orbital height ratio was used to determine symmetry between the two orbital heights. It was the distance from the supraorbital notch to the zygomaticomaxillary suture of the ipsilateral orbit over that of the contralateral orbit. The orbital width ratio was the distance from the dacryon to the zygomaticofrontal suture of the ipsilateral orbit divided by that of the contralateral orbit (Fig. 1).
Given the lack of confirmed normality within the data set, we used nonparametric methods (Wilcoxon signed ranks tests) to compare cases to age-matched controls. We compared cases at the different time points (time 0, time 1, and time 2) in the same fashion, and used the Bonferroni procedure to adjust for multiple comparisons. We calculated adjusted associations between the predictor variables and bandeau ratio using a multiple logistic regression model. Statistical significance was defined as p < 0.05 (two-tailed).
Forty-three (10 boys and 33 girls) of the 48 unicoronal synostosis cases in the study period met inclusion criteria of having all three computed tomographic scans performed. The study group’s mean age was 0.63 ± 0.22 year (range, 0.20 to 1.14 years) at time 0, 0.88 ± 0.16 year (range, 0.66 to 1.39 years) at time 1, and 3.07 ± 0.52 years at time 2 (range, 2.25 to 5.38 years), based on family follow-up preference and availability.
Thirty-six controls (15 boys and 21 girls) were separated to match the three study time points: time 0 controls (n = 12; mean age, 0.55 ± 0.15 year; range, 0.35 to 0.76 year), time 1 controls (n = 12; mean age, 1.02 ± 0.15 years; range, 0.80 to 1.23 years), and time 2 controls (n = 12; mean age, 2.85 ± 0.36 years; range, 2.55 to 3.42 years).
Bandeau ratios are listed in Table 1. The median unicoronal synostosis bandeau ratio was 0.76 (range, 0.59 to 0.97) at time 0 (ipsilateral retrusion), 1.13 (range, 0.89 to 1.34) at time 1 (slight overcorrection), and 1.01 (range, 0.82 to 1.22) at time 2 (near symmetry). There was a significant difference between all three time points (p < 0.001) (Fig. 3). The bandeau ratio at time 0 was lower, and the bandeau ratio at time 1 was greater than in age-comparable controls (p < 0.001). The bandeau ratio at time 2 was not different from that in age-comparable controls (p = 0.51).
At time 0, nine subjects had severe deformities (bandeau ratio < 0.7), 28 subjects had moderate deformities (bandeau ratio of 0.7 to 0.85), and six subjects had mild deformities (bandeau ratio > 0.85). At time 1, six subjects were undercorrected (bandeau ratio < 1.0), 26 patients were overcorrected (bandeau ratio of 1 to 1.2), and 11 patients were substantially overcorrected (bandeau ratio > 1.2). Fifty-five percent of the cases with severe presentation were overcorrected after surgery, compared with 94 percent of the moderate and mild presentations (Table 2). When compared to patients with severe deformities (bandeau ratio < 0.7) and adjusting for the effects of age at the time of surgery and operating surgeon, patients with mild to moderate deformities (bandeau ratio > 0.7) at time 0 were significantly more likely to have an overcorrection or substantial overcorrection at time 1 (OR, 4.2; 95 percent CI, 1.1 to 16.7; p = 0.04).
At time 2, 24 of the 43 subjects had a desired symmetric or overcorrected result (bandeau ratio > 1.0). Five of the six patients with mild asymmetry at presentation, 17 of 28 with moderate asymmetry, and two of nine with severe asymmetry had a symmetric or overcorrected result 2 years after surgery (p = 0.04). After adjusting for the effects of patient age at the time of surgery and surgeon, the odds of a desired symmetric or overcorrected result at time 2 (bandeau ratio > 1.0) were associated with a less severe deformity at time 0 (OR, 4.2; 95 percent CI, 1.1 to 16.7; p = 0.04) (Table 3).
Twelve of the 26 patients corrected and 10 of 11 of those overcorrected at surgery (time 1) had symmetry or overcorrection (bandeau ratio > 1) at time 2. In comparison, two of the six patients who were undercorrected at time 1 achieved this goal. After adjusting for the effects of patient age at the time of surgery and surgeon, patients who had symmetry or overcorrection at time 1 were four times more likely to have a symmetric or overcorrected result at time 2 (OR, 4.2; 95 percent CI, 1.3 to 18.0; p = 0.02) (Table 3).
The average bandeau ratio change from time 1 to time 2 was −9.1 ± 9.4 percent (range, −27.9 to 15.9) and was impacted by the degree of symmetry immediately after surgery (time 1). Undercorrected bandeaus (bandeau ratio at time 1 < 1) had a mean change of 3.0 ± 3.9 percent (i.e., no net relapse; p < 0.003), whereas those overcorrected (bandeau ratio at time 1 of 1 to 1.2) and those who were significantly overcorrected (bandeau ratio at time 1 > 1.2) at surgery had higher changes in bandeau ratio in the subsequent 2 years (−10.5 ± 9.7 percent and −12.3 ± 5.2 percent, respectively).
Skull Base and Midface Twists
Summary statistics for skull base twist are listed in Table 1. The skull-base twist was significantly different from time 0 to time 2. The unicoronal synostosis patients had significantly greater skull base twist at all three time points compared with controls (p < 0.0001).
Summary statistics for midface twist are listed in Table 1. The unicoronal synostosis patients had midface twist toward the contralateral side at time 0, which was unchanged (p = 1.00) at time 1 but improved (p < 0.01) at time 2. The unicoronal synostosis midface twist was significantly greater than controls at all three time points (p < 0.001). Preoperatively, midface twist was categorized as mild [five of 43 (11.6 percent)], moderate [27 of 43 (62.8 percent)], or severe [nine of 43 (20.9 percent)]. At time 2, midface twist improved significantly (p = 0.03), with 29 patients (67.4 percent) having mild twists, 10 patients (23.3 percent) having moderate twists, and four patients (9.3 percent) having severe twists (Table 4).
Orbital Height Ratio
Summary statistics for orbital height ratios are listed in Table 5. Compared to control patients, the unicoronal synostosis patients had taller orbits on the affected side at time 0. This asymmetry was initially corrected by surgery, as the time 1 orbital height ratio decreased (p < 0.001) to 1.00 ± 0.07. At time 2, the ipsilateral relative orbit height increased but remained lower than before surgery (1.02 ± 0.07; range, 0.88 to 1.22; p = 0.002). The control patients had symmetric orbital heights in all three age groups (orbital height ratio, 0.99 to 1.00). Compared with controls, the unicoronal synostosis patients had significant orbital height asymmetry preoperatively (p < 0.001) but no difference in height ratios after surgery (p > 0.10).
Cases with an onlay bone graft on the affected orbital rim (n = 33) had an orbital height ratio of 1.09 ± 0.04 at time 0, 1.01 ± 0.08 at time 1, and 1.04 ± 0.07 at time 2. Cases with no graft and orbital lowering by lateral rim excision (n = 10) had an orbital height ratio of 1.03 ± 0.07 at time 0, 0.99 ± 0.07 at time 1, and 0.99 ± 0.04 at time 2. We did not perform statistical comparison of these two groups because the use of graft was surgeon preference and was therefore an unavoidable confounder.
Orbit Width Ratio
Summary statistics for orbital width ratios are listed in Table 5. The width of the ipsilateral orbit was narrower than that of the contralateral orbit at time 0, and remained unchanged at time 1 and time 2 (p = 1.00). The unicoronal synostosis orbital widths were significantly more asymmetric at all three time points compared with the age-comparable controls (p < 0.001).
We measured computed tomographic scan bone changes before, immediately after, and 2 years after fronto-orbital advancement for the treatment of unicoronal synostosis. Our analysis supports the clinical observation that long-term forehead symmetry is best achieved by overcorrection of the bandeau at the time of surgery. The degree of presurgery bandeau asymmetry also had a strong association with the 2-year result, with severe deformities being more likely to have residual asymmetry 2 years postoperatively. We identified a mild relapse toward the preoperative state in both orbital height correction and bandeau advancement. With our current approach, the narrow ipsilateral orbit was not corrected by surgery, and offers opportunity for evolution in technique.
We used the bandeau ratio to analyze lateral and anterior symmetry of the brow one-sixth of the forehead height above the orbits and in the critical radian area of 30 to 60 degrees from the midline. Although this single variable measures the axial region of greatest change before and after surgery, it must be emphasized that it does not represent the overall aesthetic result, which is a gestalt of multiple variables. We observed that patients who had an overcorrection of the bandeau ratio at the time of surgery were five times more likely to have a symmetric or overcorrected result 2 years later. The degree of overcorrection, however, did affect the amount of brow relapse that occurred in the years after surgery, with greater advancements being associated with larger percentage reversion toward the preoperative deformity. Ten of the 11 patients who achieved a greater than 20 percent overcorrection of the bandeau ratio had a bandeau ratio greater than 1.0 at 2 years, despite an average relapse of 12 percent.
Despite our surgical goal of brow overcorrection, six of our 43 patients were still undercorrected (bandeau ratio < 1.0) immediately after surgery. One of the greatest influences of an inability to overcorrect may be the severity of the deformity at presentation. We corrected or overcorrected 94 percent of mild and moderate deformities, but corrected 54 percent of the severe presentations, and did not substantially overcorrect any of them. This would suggest that there is a soft-tissue threshold in severe cases that challenges overcorrection. The final position of the brow at time 2 was likely determined not only by the degree of bone overcorrection, but also by the degree to which the underlying brain expanded to support the graft during remodeling. This would explain our finding that the greater the advancement of the brow at surgery, the greater the percentage relapse over time.
Other authors share our concern for progressive reversion toward the untreated deformity in unicoronal synostosis years after surgery. Martini et al. noted complete relapse of the frontoparietal angle change in two of their four unicoronal synostosis patients at 1 year.14 Fearon et al. observed a drop in head circumference in unicoronal synostosis patients from a z value of 0.726 after surgery to 0.03 an average of 4 years later.8 Becker et al. also showed a progressive increasing asymmetry in six patients followed up to skeletal maturity, but that the final result remained better than untreated controls.15 Taylor et al. performed a retrospective chart review of physical findings in unicoronal synostosis patients treated with fronto-orbital advancement.9 They found that 75 percent of patients with at least 1-year follow-up were noted to have brow retrusion, which was more likely in patients who had 5 or more years of follow-up. Based on these studies and our own data, we propose that when treating unicoronal synostosis with fronto-orbital advancement, the goal should not be symmetry at 2 years after surgery, but should actually be overcorrection at this relatively young age. Our results suggest that at least a 20 percent overcorrection provides the best opportunity to achieve early overcorrection, but longer term follow-up will be required to determine whether this is successful in achieving symmetry at skeletal maturity.
After controlling for patient age and surgeon, the severity of bandeau deformity at presentation was also associated with achieving the goal of symmetry or overcorrection at 2 years. Patients with severe brow retrusion at presentation were four times more likely to have persistent asymmetry at follow-up. Of the seven patients with a bandeau ratio less than 0.7, only two had a bandeau ratio greater than 1.0 at 2 years after surgery. This observation can be helpful in presurgical counseling of families on the likelihood of persistent brow asymmetry, and the need for techniques to allow overcorrection of these severe cases.
The anterior skull base twist has been reported, but no prior studies have evaluated how this twist changes after surgery. Showalter et al. measured a mean 11.3-degree twist in their cohort of 10 patients with isolated unicoronal synostosis.4 Our results show an 11.4-degree twist preoperatively, which remained unchanged at 11.1 degrees after surgery. After 2 years, however, the anterior skull base twist statistically improved to 9.8 degrees. Although this is statistically significant, it is unclear whether this improvement is clinically relevant. It does, however, suggest that there are some secondary changes to the unicoronal synostosis anterior skull base morphology within the first 2 years after fronto-orbital advancement. We did not have a nonoperated unicoronal synostosis group as a comparison to determine whether fronto-orbital advancement “unlocks” the constraint on the anterior skull base, or whether this change would occur with time regardless of surgery.
The midface twist also significantly improved at the 2-year follow-up, from 8.1 degrees to 5.3 degrees, despite not having an improvement immediately after surgery. This was consistent with the subjective observations by McCarthy et al. of 25 of 32 patients having resolution of midface twist at long term follow-up after fronto-orbital advancement.3 Oh et al. measured midface twist on photographs of 15 patients greater than 10 years after fronto-orbital advancement for unicoronal synostosis during infancy.16 Although their measurements were soft-tissue based, they measured a midface twist of 3.4 degrees in these older children, which is lower than the 5.3 degrees we measured in our 2-year group. This may suggest that midface twist continues to improve beyond our 2-year time point.
The orbital asymmetry in patients with unicoronal synostosis results from dysmorphology in both orbits.17 On clinical examination of 33 unicoronal synostosis patients after fronto-orbital advancement by McCarthy et al., 12 had complete resolution, 14 had improvement, and six had no change of their orbital dysplasia at a mean 4.3-year follow-up.3 In our 43 unicoronal synostosis patients, the orbital height ratio was corrected from 1.07 to 0.99 with surgery. However, a milder but still statistically significant harlequin deformity recurred at 2 years with an orbital height ratio of 1.03. We did not identify a clear advantage of onlay grafting in our series. We attempted to create symmetry of the orbits immediately after surgery, but this mild recurrence toward the preoperative deformity prompts us to consider slight overcorrection of orbital height in future patients.
Although the orbital height was corrected with surgery, the orbital width was not. Steinbacher et al. demonstrated that patients with unicoronal synostosis had narrower ipsilateral orbits compared with controls, and Derderian et al. observed that the ipsilateral orbital width ratios remained unchanged after surgery.18,19 Our data were consistent with these findings and support surgical widening of the ipsilateral orbit. One option is widening the ipsilateral orbit by outfracturing the lateral orbital rim portion of the bandeau at the zygomaticofrontal suture and fixating this piece more laterally onto the zygoma. Alternatively, the entire bandeau could be shifted slightly toward the ipsilateral side during fixation to simultaneously adjust both orbits. Jackson et al. proposed horizontal expansion of the ipsilateral orbit with an interposition bone graft at the superior orbital rim and reported subjectively good outcomes.20
This quantitative analysis adds to the evidence-based presurgical counseling we can offer parents regarding expected outcomes based on presenting deformity. In addition, our findings confirm opportunities for continued improvement in surgical technique. We have adjusted our goals of fronto-orbital advancement to overcorrect the bandeau ratio 20 percent beyond symmetry, slightly overcorrect the harlequin deformity, and shift the bandeau toward the ipsilateral side to widen the ipsilateral orbit and narrow the contralateral orbit.
The authors would like to thank Ezgi Mercan Keremoglu, Ph.D., for assistance in preparing the figure images, and they would like to thank Robyn Randall, B.S., for assistance to David Y. Khechoyan, M.D., in contribution to the article.
1. Lajeunie E, Le Merrer M, Bonaïti-Pellie C, Marchac D, Renier D. Genetic study of nonsyndromic coronal craniosynostosis. Am J Med Genet. 1995;55:500–504.
2. Slater BJ, Lenton KA, Kwan MD, Gupta DM, Wan DC, Longaker MT. Cranial sutures: A brief review. Plast Reconstr Surg. 2008;121:170e–178e.
3. McCarthy JG, Glasberg SB, Cutting CB, et al. Twenty-year experience with early surgery for craniosynostosis: I. Isolated craniofacial synostosis. Results and unsolved problems. Plast Reconstr Surg. 1995;96:272–283.
4. Showalter BM, David LR, Argenta LC, Thompson JT. Influence of frontosphenoidal suture synostosis on skull dysmorphology in unicoronal suture synostosis. J Craniofac Surg. 2012;23:1709–1712.
5. Ploplys EA, Hopper RA, Muzaffar AR, et al. Comparison of computed tomographic imaging measurements with clinical findings in children with unilateral lambdoid synostosis. Plast Reconstr Surg. 2009;123:300–309.
6. Öwall L, Darvann TA, Larsen P, et al. Facial asymmetry in children with unicoronal synostosis who have undergone craniofacial reconstruction in infancy. Cleft Palate Craniofac J. 2016;53:385–393.
7. Xu W, Gerety PA, Li J, Bartlett SP, Taylor JA. Frontal sinus volume and frontal bar position are associated in unicoronal craniosynostosis. Plast Reconstr Surg. 2016;139:965e–974e.
8. Fearon JA, Ruotolo RA, Kolar JC. Single sutural craniosynostoses: Surgical outcomes and long-term growth. Plast Reconstr Surg. 2009;123:635–642.
9. Taylor JA, Paliga JT, Wes AM, et al. A critical evaluation of long-term aesthetic outcomes of fronto-orbital advancement and cranial vault remodeling in nonsyndromic unicoronal craniosynostosis. Plast Reconstr Surg. 2015;135:220–231.
10. Morton RP, Reynolds RM, Ramakrishna R, et al. Low-dose head computed tomography in children: A single institutional experience in pediatric radiation risk reduction. Clinical article. J Neurosurg Pediatr. 2013;12:406–410.
11. Grant JH III, Roberts TS, Loeser JD, Gruss JS. Onlay bone graft augmentation for refined correction of coronal synostosis. Cleft Palate Craniofac J. 2002;39:546–554.
12. Montoya JC, Eckel LJ, DeLone DR, et al. Low-dose CT for craniosynostosis: Preserving diagnostic benefit with substantial radiation dose reduction. AJNR Am J Neuroradiol. 2017;38:672–677.
13. Ezaldein HH, Metzler P, Persing JA, Steinbacher DM. Three-dimensional orbital dysmorphology in metopic synostosis. J Plast Reconstr Aesthet Surg. 2014;67:900–905.
14. Martini M, Schulz M, Rohrig A, et al. A 3D morphometric follow-up analysis after frontoorbital advancement in non-syndromic craniosynostosis. J Craniomaxillofac Surg. 2015;43:1428–1437.
15. Becker DB, Fundakowski CE, Govier DP, Deleon VB, Marsh JL, Kane AA. Long-term osseous morphologic outcome of surgically treated unilateral coronal craniosynostosis. Plast Reconstr Surg. 2006;117:929–935.
16. Oh AK, Wong J, Ohta E, Rogers GF, Deutsch CK, Mulliken JB. Facial asymmetry in unilateral coronal synostosis: Long-term results after fronto-orbital advancement. Plast Reconstr Surg. 2008;121:545–562.
17. Beckett JS, Persing JA, Steinbacher DM. Bilateral orbital dysmorphology in unicoronal synostosis. Plast Reconstr Surg. 2013;131:125–130.
18. Steinbacher DM, Wink J, Bartlett SP. Temporal hollowing following surgical correction of unicoronal synostosis. Plast Reconstr Surg. 2011;128:231–240.
19. Derderian CA, Wink JD, Cucchiara A, Taylor JA, Bartlett SP. The temporal region in unilateral coronal craniosynostosis: A volumetric study of short- and long-term changes after fronto-orbital advancement. Plast Reconstr Surg. 2014;134:83–91.
©2019American Society of Plastic Surgeons
20. Jackson IT, Costanzo C, Marsh WR, Adham M. Orbital expansion in plagiocephaly. Br J Plast Surg. 1988;41:16–19.