Bony orbital maldevelopment after enucleation

Abstract

One common belief in ophthalmology is that enucleation at an early age will result in bony orbital maldevelopment and facial asymmetry. However, the age range in which enucleation is associated with risk of orbital maldevelopment and the extent of asymmetry remains controversial. In this study, patients who had undergone unilateral enucleation at different ages without orbital implantation were analysed to investigate bony orbital development after enucleation. A total of 87 Chinese adult patients were included. Their bony orbital volume and orbital aditus area were measured using three-dimensional reconstructive models based on patients' computer tomography scans. The ratio of the parameter values of the affected orbit to the unaffected orbit was calculated and described as the orbital symmetry index. The results showed that the bony orbit grew until approximately 18 years old. Enucleation after that age did not affect the orbit, whereas enucleation before that age led to significant orbital maldevelopment. The relative reduction ranged up to 20% in orbital volume and 17% in the orbital aditus area. The extent of orbital maldevelopment was correlated to the age of enucleation. The symmetry index of orbital volume = −0.0003x² + 0.0159x + 0.8112 (x = the age of enucleation). The symmetry index of the orbital aditus area = −0.0002x² + 0.0119x + 0.8504 (x = the age of enucleation). The regression formulae were used to predict the severity of orbital asymmetry after unilateral enucleation, and evaluate the necessity and efficacy of interventions following enucleation.

Introduction

From birth to adulthood, the orbit enlarges with the growth of the entire skull. The normal bony orbital structure has left to right symmetry during the growth period and in the maturation stage (Forbes et al. 1985; McGurk et al. 1993; Furuta, 2001; Bentley et al. 2002; Acer et al. 2009). However, the symmetry of the orbital structure is known to be broken after unilateral enucleation in some cases, which presents as a smaller bony orbital rim along with soft-tissue changes on the affected side. These cases are mostly related to enucleation at an early age (Apt & Isenberg, 1973; Osborne et al. 1974; Peylan-Ramu et al. 2001; Yago & Furuta, 2001). Kennedy confirmed this detrimental influence of early enucleation on orbital growth through experiments on animal models in 1964, and estimated that the orbital volume reduction after enucleation could reach a magnitude of 35–50% (Kennedy, 1964). In clinical cases, the difference between the affected and unaffected sides does not present soon after enucleation because the bony orbital maldevelopment takes years to emerge. In the literature, the age of enucleation and the time interval between enucleation and examination are thought to be influencing factors, but their effects have not yet been ascertained (Kaste et al. 1997; Hintschich et al. 2001).

With respect to the evaluation of bony orbital size, the orbital volume was the only parameter generally used in previous studies (Kennedy, 1964; Kaste et al. 1997; Furuta, 2001; Hintschich et al. 2001; Peylan-Ramu et al. 2001; Yago & Furuta, 2001). Whereas the bony orbital volume describes the size of the orbital cavity but not the dimensions of the orbital rim, other orbital parameters should be introduced to provide a more direct representation of the size of the orbital rim, which has equal cosmetic significance on overall appearance.

To address these issues, a three-dimensional (3D) reconstruction imaging technique was used in this study to measure the orbital volume and orbital aditus area in unilateral enucleation patients without orbital implantation (Deveci et al. 2000; Park et al. 2006; Regensburg et al. 2008; Ji et al. 2010). Two matters were investigated: first, the age range in which enucleation introduces risk for bony orbital maldevelopment in later years; and second, the extent to which enucleation may affect the orbit symmetry.

Subjects and methods

Subjects

Adult patients referred to Shanghai 9th People's Hospital from 1 January 2009 to 30 June 2013 for a consultation regarding second-stage orbital implantation were retrospectively studied. All of the patients had undergone unilateral enucleation previously but did not receive one-stage orbital implantation at the time of enucleation, mainly because the cost of the implant and implantation surgery was too costly or the implantation surgery was not proposed by their ophthalmologists. Orbital implantation (including one-stage implantation and second-stage implantation) is recommended, but it is not a customary intervention after enucleation in China. The patients who had congenital facial malformation, orbital fracture, craniofacial radiotherapy or any endocrine disease that may affect the orbits were excluded from this study. The patients whose time interval between enucleation and orbital computer tomography (CT) examination was < 1 year were not recruited. The patients whose digital CT data were not available in this hospital were excluded. The research adhered to the tenets of the Declaration of Helsinki. The Medical Ethics Committee of Shanghai 9th People's Hospital did not require informed consent for this retrospective study.

Image analysis

Computer tomographic results for all subjects were analysed in this study. Craniofacial scans were obtained with multi-slice CT (GE LightSpeed 64, Milwaukee, WI, USA) using high-resolution contiguous sections in an axial plane for 3D reconstruction with the following protocol: tube voltage, 120 KVp; tube current, 100–140 mA; slice thickness, 1.25 mm; field of view, 25 × 25 cm; and matrix, 512 × 512. Image analysis was performed using SimMed software (Shanghai Jiantong University, China; Fig.1). Three-dimensional images of the orbito-facial region were reconstructed and displayed on the basis of the original axial section images. Manual segmentation was manipulated in a soft-tissue window to define the boundary of the orbital contents. The optic foramen and four orbital walls were defined as the borders of the bony orbit. A simulated surface was defined to cover the orbital aditus with the orbital rim. This surface served as the anterior border of the bony orbit. Two orbital parameters of the study, i.e. the bony orbital volume and orbital aditus area, were measured automatically using software tools.

Fig 1.

The process of CT image analysis and 3D model reconstruction. (A) The boundary of the orbit was defined in each slice by manual segmentation in a soft-tissue window. (B) 3D model of the orbito-facial region was built on the basis of CT slices by the software. The orbital rim was drawn directly on a 3D model (the red curve). A simulated surface was generated to cover the orbital aditus (the blue region) with the orbital rim. The orbital aditus area was measured by software tools. (C) Models of bilateral orbital cavity (in blue). The volume of the models was calculated with software tools.

Statistical analysis

The statistical analysis was performed with Excel 2007 (Microsoft Office, Microsoft, USA) and spss 14.0 (SPSS, Chicago, IL, USA). For each parameter, the ratio of the parameter value of the affected orbit to the unaffected was calculated and described as the symmetry index. A symmetry index value < 1.00 meant that the affected orbit was smaller than the unaffected side at the level of this parameter. Symmetry indices were plotted in scatter diagrams with increasing age of enucleation. Regression curves were generated by software to best simulate the variation tendency of the symmetry indices with the age of enucleation. Regression formulae were therefore fitted. The age range in which the regression curve was below 1.00 meant that enucleation in that age range would cause orbital maldevelopment.

According to the regression curve, the orbital symmetry in certain age groups was estimated by correlation analysis or a paired samples t-test.

Results

Summary of subjects

Eighty-seven adult Chinese patients (47 males, 40 females) were included in this study. The age at the CT scan examination in the hospital ranged from 24 years to 58 years. The age of enucleation ranged from 1 year to 40 years. The time interval between enucleation and CT examination ranged from 2 years to 40 years. Of these patients, 42 subjects had undergone enucleation for the right eye, and 45 subjects had undergone enucleation for the left eye. Twenty-two patients had undergone enucleation to treat unilateral ocular tumours, 55 patients for ocular trauma and 10 patients for severe endophthalmitis. All the unaffected eyes of these patients presented with normal appearance and visual acuity. Some patients wore an ocular prosthesis after enucleation, but their wearing time and replacement of the prosthesis were not documented.

3D morphometric findings

Three-dimensional models of the orbito-facial region were reconstructed based on CT data for each subject. Compared with those who had enucleation in adulthood, the subjects with early enucleation presented with more differences between the affected and unaffected orbits, which was apparent on 3D images and also proven by the values of orbital symmetry indices (Fig.2).

Fig 2.

Comparison of 3D morphometric images between an early enucleation example and a late enucleation example. Early enucleation example: male subject, enucleation of the right eye at 2 years old. (A) Orbital aditus area was 10.34 cm² on the affected side (right side, in blue) and 11.88 cm² on the unaffected side (left side, in green). The symmetry index of the orbital aditus area was 0.87. (B) The bony orbital volume was 22.15 mL on the affected side (right side, in blue) and 26.69 mL on the unaffected side (left side, in green). The symmetry index of the bony orbital volume was 0.83. Late enucleation example: male subject, enucleation of the left eye at 25 years old. (C) The orbital aditus area was 12.22 cm² on the affected side (left side, in blue) and 11.98 cm² on the unaffected side (right side, in green). The symmetry index of the orbital aditus area was 1.02. (D) The bony orbital volume was 26.85 mL on the affected side (left side, in blue) and 26.58 mL on the unaffected side (right side, in green). The symmetry index of the bony orbital volume was 1.01.

Statistical analysis

Scatter diagrams (Fig.3) were created to show the tendency of the symmetry indices that varied with the age of enucleation. The regression curves of symmetry indices for both the bony orbital volume and orbital aditus area increased with increasing age of enucleation, reached a value of 1.00 and then remained at a plateau of approximately 1.00. Both the male subjects and female subjects followed the same trend. Regression formulae were then fitted for all subjects. The symmetry index of orbital volume = −0.0003x² + 0.0159x + 0.8112 (x = the age of enucleation, R² = 0.9407). The curve crossed the value 1.00 at 17.96 years old (Fig.3A). The symmetry index of the orbital aditus area = −0.0002x² + 0.0119x + 0.8504 (x = the age of enucleation, R² = 0.9037). The curve crossed the value 1.00 at 18.04 years old (Fig.3B).

Fig 3.

Scatter diagram of the symmetry index of the bony orbital volume (A) and orbital aditus area (B) varying with the age of enucleation in male subjects (□) and in female subjects (▵). The regression curves of all subjects were fitted. (A) The symmetry index of the orbital volume = −0.0003x² + 0.0159x + 0.8112 (x = the age of enucleation, R² = 0.9407). The curve crossed the value 1.00 at 17.96 years old. (B) The symmetry index of the orbital aditus area = −0.0002x² + 0.0119x + 0.8504 (x = the age of enucleation, R² = 0.9037). The curve crossed the value 1.00 at the 18.04 years old.

As both curves crossed the value 1.00 at about age 18 years old, the subjects were then divided into two groups:

1. group 1, age of enucleation < 18 years;

2. group 2, age of enucleation equal to and greater than 18 years.

The summary of orbital parameter values is listed separately for each group. In group 1, the symmetry index ranged from 0.80 to 1.02 for the bony orbital volume and from 0.83 to 1.02 for the orbital aditus area. Significant positive correlations were shown between orbital symmetry indices and the age of enucleation in both male and female subjects. In group 2, paired sample t-tests were carried out and showed no significant morphological differences between the affected and unaffected orbits (Table1).

Table 1.

Orbital parameter values divided into two groups by the age at which enucleation was performed

Age of enucleation (years)	Number of subjects	Parameters	Gender	Affected orbit (minimum–maximum)	Unaffected orbit (minimum–maximum)	Symmetry index (minimum–maximum)	Correlation analysis of symmetry index and age of enucleation*
< 18	53 (29 males, 24 females)	Bony orbital volume (mL)	M	19.70–26.65	23.54–27.92	0.81–1.01	R = 0.964 P < 0.001
			F	16.75–24.88	20.27–25.12	0.80–1.02	R = 0.963 P < 0.001
		Orbital aditus area (cm2)	M	9.49–12.73	10.41–13.54	0.83–1.01	R = 0.956 P < 0.001
			F	8.80–12.77	9.88–12.77	0.84–1.02	R = 0.949 P < 0.001

				Affected orbit (mean ± deviation)	Unaffected orbit (mean ± deviation)	Symmetry index (mean ± deviation)	Paired samples t-test**
≥ 18	34 (18 males, 16 females)	Bony orbital volume (mL)	M	26.41 ± 1.30	26.34 ± 1.27	1.00 ± 0.01	t = −0.954 P = 0.354
			F	23.32 ± 1.57	23.20 ± 1.80	1.01 ± 0.01	t = 1.859 P = 0.083
		Orbital aditus area (cm2)	M	11.54 ± 0.84	11.53 ± 0.76	1.00 ± 0.01	t = −0.438 P = 0.667
			F	11.16 ± 0.85	11.11 ± 0.83	1.01 ± 0.01	t = 1.136 P = 0.194

^*In the correlation analysis of the orbital symmetry index and age of enucleation, R represents the Spearman correlation coefficients. A high R-value and P-value < 0.05 indicate a significant correlation between the orbital symmetry index and the age of enucleation.

^**In a paired samples t-test, a P-value > 0.05 indicates that the orbital parameter values of the affected and unaffected sides had no significant difference.

Discussion

Enucleation is the removal of the eyeball from the orbit, involving separation of all connections between the eyeball and the patient. The three most common indications for enucleation are intraocular malignancy, a blind painful eye, and prevention of sympathetic ophthalmia. Other indications include phthisis and improvement of cosmesis (Moshfeghi et al. 2000). In the current study, ocular tumour, ocular trauma and severe endophthalmitis were the three causes of unilateral enucleation. As the eyeball is the major intra-orbital content providing persistent intra-orbital pressure, especially in the eyeball growth period, to stimulate the simultaneous orbital growth (Weiss et al. 1989; Farkas et al. 1992; Kaste et al. 1997), one common belief in ophthalmology is that the orbital growth needs to be checked for eventual orbital maldevelopment and facial asymmetry in the cases of enucleation at an early age (Kennedy, 1964; Apt & Isenberg, 1973; Osborne et al. 1974). In this study, adult patients were enrolled who had undergone unilateral enucleation without orbital implantation as subjects to retrospectively observe bony orbital development following enucleation without intervention. As a morphometric orbital parameter, orbital volume is generally used and even has been the only parameter in previous studies on orbital growth and malformation (Kennedy, 1964; Kaste et al. 1997; Furuta, 2001; Hintschich et al. 2001; Peylan-Ramu et al. 2001; Yago & Furuta, 2001; Lin & Liao, 2011). It can well represent the size of the orbital cavity but not the orbital rim. Therefore, another morphometric parameter, i.e. the orbital aditus area, was introduced in this study to describe the dimension of the orbital rim.

Although the cause of orbital maldevelopment is mostly due to enucleation in childhood, it has remained controversial whether an enucleation in adulthood would result in bony orbital shrinkage (Hintschich et al. 2001). The results of this study showed that the impact of enucleation on orbital symmetry presented an inflection point at about 18 years old. Statistic asymmetry of the two orbits was found in the cases of unilateral enucleation before 18 years old, but not at ages equal or greater than 18 years. This finding was similar to some previous studies and suggested that the age range in which an enucleation can affect the orbit size is limited to near 18 years old (Osborne et al. 1974; Chau et al. 2004). Therefore, it is reasonable to presume that the orbit enters a stable status after the age of approximately 18 years, and the bony orbit will not shrink if an enucleation is performed after that age. The intervention aimed at enlarging the affected orbit is therefore not necessary in the cases of enucleation in adulthood.

Moreover, this finding is helpful for the study of orbital growth. The asymmetry of orbital size in this study was not caused by shrinkage of the affected side, but was the sign of different orbital growth rates between the affected and unaffected sides. Therefore, the age of approximately 18 years can be considered as the end point of orbital growth. Compared with the data in the literature, the age of the end point in the current study was later than some previous studies (Farkas et al. 1992; Furuta, 2001; Bentley et al. 2002), but close to the outcome of the study by Chau et al. (2004) on Chinese subjects from Hong Kong. The difference may be due to racial differences and the use of different statistical methods. In previous studies, the orbital growth curve was generally drawn according to the scatter diagram of the absolute orbital parameter values acquired from individuals of different ages (Furuta, 2001; Bentley et al. 2002; Chau et al. 2004). The individual differences in orbital size in the same age group may have directly altered the outcome. The current study used the symmetry of orbital structures as a baseline and the relative values instead of absolute values to avoid individual deviations.

The extent to which enucleation affects the orbital symmetry varies widely in the literature. Kennedy calculated the decreased orbital volume to be 20–30% according to human roentgenograms, and he empirically determined that the volume reduction after enucleation in humans can reach a magnitude of 35–50% from the animal experiment results (Kennedy, 1964). Yago found the two largest differences in the orbital volume between the two sides were 28% and 37%, respectively (Yago & Furuta, 2001). In contrast, Hintschich found that the greatest volume reduction was 14.5% in his study; clinically, none of the patients showed signs of facial asymmetry (Hintschich et al. 2001). In the current study, for the subjects who had enucleation before 18 years old, the affected orbits had a relative reduction ranging to 20% in orbital volume and 17% in the orbital aditus area.

The age of enucleation and the time interval between enucleation and examination were considered to be associated with the extent of orbital maldevelopment. A previous study showed that the earlier the enucleation is performed and with more time after enucleation, the greater the difference between the two sides (Apt & Isenberg, 1973; Osborne et al. 1974; Hintschich et al. 2001; Peylan-Ramu et al. 2001; Yago & Furuta, 2001). In this study, the age of enucleation was shown to be correlated to the extent of orbital maldevelopment. Formulae were produced by regression analysis to simulate the trends in the orbital volume and orbital aditus area. These formulae are valuable in clinical practice, as the patient's prognostic orbital symmetry in adulthood can therefore be predicted with the known age of enucleation. This study also revealed that orbital growth reached an end stage at about 18 years old, which suggests that after the age of 18 years the difference between the two orbital sides would not enlarge.

As reported in a previous study by Bentley et al. (2002), the normal orbit volume increases from the first few months of life to 5 years old by a factor of 1.7 in boys and 1.8 in girls. By the time the child has reached 5 years old, the orbital volume has reached 77% of the volume observed at 15 years old in both sexes. The current results were compared with the previous study and showed that enucleation during the orbit growth period reduced the growth rate of the affected side but did not totally arrest its growth. According to the current formulae, even with enucleation at the age of 1 year and without intervention, the affected orbit can still grow to 82.68% in orbital volume and 86.21% in the orbital aditus area of the unaffected orbit when reaching adulthood. As the orbital cavity is formed by seven bones, i.e. the maxilla, palatine, frontal, sphenoid, zygomatic, ethmoid and lacrimal bones, the orbit still undergoes a certain amount of growth over the time course after enucleation, along with whole craniofacial development, even without the stimulation of intra-orbital pressure. This amount of growth should be taken into consideration, especially in the evaluation of intervention to prevent orbital growth retardation. A simple comparison of orbital size before and after intervention cannot prove the effectiveness of the intervention, such as the orbital implant and expander (Mazzoli et al. 2004; Tse et al. 2007), as the enucleated orbit also grows itself, even without intervention. The formulae in this study may provide a baseline evaluation standard for such cases.

Some subjects involved in this study wore prosthesis after enucleation. The size and material of the prosthesis were not noted in their follow-up documents. Theoretically, a prosthesis serves to replace the orbital volume as the implant, but actually it was rarely considered as a major intervention to prevent orbital growth retardation, mainly for the reasons below. First, the prosthesis is detachable. The wear and replacement of the prosthesis depends on the patient. Secondly, the prosthesis is placed in a conjunctival sac and lays its weight mainly on the lower eyelid. A large prosthesis may induce ectropion or eyelid laxity. Therefore, an ocular prosthesis cannot provide a volume more than 4.2 mL, and its ideal volume has been suggested to be 2.2 mL (Kaltreider, 2000; Yago & Furuta, 2001). In this study, there was no evidence to suggest that orbital growth retardation was prevented by an ocular prosthesis. However, further research is still necessary to clarify whether a specially designed prosthesis or regular replacement of a larger prosthesis could potentially promote the growth of the orbital rim. If so, it would offer a more cosmetic option than an expander of the orbital cavity.

Conclusions

This study investigated the extent to which enucleation at different ages affects orbit development. Orbital morphometric parameters were measured with a 3D reconstructive method to quantitatively describe orbital symmetry. The results showed that the bony orbit grew until the age of approximately 18 years. Enucleation before that age led to significant orbital maldevelopment, the extent of which was correlated to the age of enucleation. However, enucleation after that age did not affect the orbit. The regression formulae generated in this study could help to predict the severity of orbital asymmetry after unilateral enucleation and evaluate the necessity and efficacy of interventions following enucleation.

Author contributions

YJ contributed to the conception and design of the study, acquisition of the data, analysis and interpretation of the data, drafting the article and approval of final the version. FY contributed to acquisition, analysis and interpretation of the data, drafting the article and approval of final the version. HZ contributed to the conception of the study, analysis of the data, revising the article critically for important intellectual content and approval of the final version. QX contributed to acquisition and analysis of the data, drafting the article and approval of the final version. SG contributed to the design and management of the study, revising the article critically for important intellectual content and approval of the final version. XF contributed to the conception and management of the study, revising the article critically for important intellectual content and approval of the final version.