Original Investigation
Assessment of Bony Nasolacrimal Parameters Among Asians Adeline Mei-Yen Yong*, Daniel Binhang Zhao, M.Sc†, Shuen Chao Siew, F.R.C.S.‡, Poh Sun Goh, F.R.C.R.†, Jiemin Liao, M.A.§║, and Shantha Amrith, F.R.C.S.§ *Yong Loo Lin School of Medicine, National University of Singapore, Singapore; †Department of Diagnostic Imaging, ‡Department of Otolaryngology, and §Department of Ophthalmology, National University Hospital, Singapore; and ║Singapore Eye Research Institute, Singapore
Purpose: A case series evaluating racial differences in the nasolacrimal region and quantifying the anterior lacrimal crest thickness and minimum nasolacrimal duct diameter in Asians. Methods: Facial or orbital CT scans of 90 consecutive patients were retrospectively reviewed. Evidence of lacrimal fossa tumor or trauma excluded a patient. Using 3-dimensional image software, the thickness of the anterior lacrimal crest, narrowest diameter of the nasolacrimal duct, vertical diameter of the lacrimal sac fossa, frontonasal angle, and inter-frontozygomatic suture distance were measured in axial, sagittal, and coronal planes. Results: Inter- and intraobserver correlation of a sample data proved reliability via intraclass correlation coefficient (0.706– 0.917). Southeast Asians had a wider inter-frontozygomatic suture distance than South Asian and Occidental races (p = 0.025). Vertical lacrimal fossa diameter, anterior lacrimal crest thickness, and narrowest nasolacrimal duct diameter did not differ significantly between right and left sides or among ethnic groups. Narrower nasolacrimal duct diameter was significantly associated with decreased inter-frontozygomatic suture distance (p < 0.001), namely in patients with narrower faces. The anterior lacrimal crest thickness was significantly affected by the nasal configuration and thicker in patients with more acute frontonasal angle (p = 0.026). Conclusions: There is no significant difference in nasolacrimal duct diameter among ethnic groups, which may predispose one to nasolacrimal duct obstruction. But, this is significantly associated with inter-frontozygomatic suture distance, suggesting that a wider face is associated with wider nasolacrimal duct diameter. An acute frontonasal angle predicts a thicker anterior lacrimal crest for surgical preparation during dacryocystorhinostomy. (Ophthal Plast Reconstr Surg 2014;30:322–327)
E
piphora caused by nasolacrimal duct obstruction (NLDO) is a common ophthalmologic problem and accounts for 3% of ophthalmologic clinic visits.1 Obstruction of the nasolacrimal drainage system can be congenital or acquired, which is
Accepted for publication November 26, 2013. The authors have no financial or conflicts of interest to disclose. Address correspondence and reprint requests to Shantha Amrith, F.R.C.S., Department of Ophthalmology, National University Hospital, 5 Lower Kent Ridge Road, Singapore, Singapore 119074. E-mail: [email protected] DOI: 10.1097/IOP.0000000000000101
322
further classified into primary and secondary causes.2 The most frequent causes of secondary acquired NLDO include facial surgery, trauma, neoplasm, sarcoidosis, and Wegener granulomatosis. In contrast, primary acquired NLDO is idiopathic and characterized by gradual chronic inflammation and fibrosis along the NLD, causing increased obstruction. Studies have suggested that racial and sex differences in facial skull dimensions may explain narrower NLDs and hence differences in NLDO incidence.3–6 External dacryocystorhinostomy (DCR) is a standard approach for treating complete NLDO.7–9 Endonasal DCR has advanced with new technology and instrumentation; however, it is still perceived as more challenging than external DCR.10–12 The endonasal approach involves creating an adequately sized and correctly located osteotomy for long-term patency of the nasolacrimal shunt.8,9,13 It is a common belief among oculoplastic surgeons that East Asian bones in the region of anterior lacrimal crest (ALC) are thicker than those of Caucasians.14 Furthermore, performing endonasal DCR on Southeast Asian patients with a low and broad nasal bridge is sometimes more difficult than on South Asian or Caucasian patients with a higher and narrower nasal bridge. Based on such surgical experience, the authors sought to determine the relationship between nose shape and facial anatomy. This is the first study performed to quantify such measurements that may explain the prevalence of NLDO and impact on DCR surgery in Asians. The authors also attempted to reveal the usefulness of knowing facial anatomy as a predictor of facility in endonasal DCR.
METHODS The authors performed a retrospective review of consecutive facial or orbital CT scans of 90 patients performed at the diagnostic imaging unit at the National University Hospital, Singapore, between January 1, 2010, and December 31, 2011. Evidence of tumor or trauma in the region of the lacrimal fossa excluded a patient. Institutional review board approval for the study was obtained. All scans were collected and numerated; patient identification data were rendered anonymous. CT Technique. All examinations were performed with a 64-slice CT scanner (Brilliance CT; Philips Healthcare, Cleveland, OH, U.S.A.). Contiguous axial 1-mm–thick sections parallel to the infraorbitomeatal line and coronal as well as sagittal scans perpendicular to this plane were acquired with a field of view of 180 mm, matrix size of 512 × 512, and a reconstruction kernel for bone. Measurements. Reconstructed images were loaded in the CT work station (Extended Brilliance Workspace; Philips), and measurements were obtained using calipers on a computer screen, as seen in
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Bony Nasolacrimal Parameters in Asians
FIG. 1. Axial CT image showing anterior lacrimal crest measurement, with correlation across 3 planes.
Figures 1, 2, and 3. The software allows for simultaneous viewing of scans in all 3 planes, and images can be toggled between different planes to verify any chosen point for measurement. Two observers, a medical student (first author) and a fellow in ophthalmic plastic and reconstructive surgery, carried out the measurements independently at different times on the first 23 scans to validate the measurements. Measurement Protocol. All measurements were performed in bone window. Thickness of the ALC was measured in the axial scans. Using the sagittal scan, the upper end of NLD was identified to make sure that the sac region was identified correctly. From that point, the thickest part of the ALC was identified by scrolling up (in axial scan) and anteriorly (in coronal scan; Fig. 1). Measurements were taken in the axial scan after the selected points were repeatedly checked in coronal section. The NLD was identified in the sagittal section, and by scrolling across adjacent cuts, the central axis of the canal was determined and a line was drawn along the long axis. The narrowest diameter was measured perpendicular to the line drawn. The start and the end point of the measure was the inner rim of the canal (Fig. 2). To ensure that the minimum diameter had been measured, several measurements were taken at different levels and the least recorded. Vertical diameter of the lacrimal sac fossa was measured using the 3-dimensional skull image. Reference points at the upper end of the NLD were marked. The point was verified in all 3 types of scans. A line was drawn from this point to where the frontomaxillary suture and the fossa intersected. The measurement of this line gave the vertical diameter of the lacrimal sac.
Frontonasal angle (FNA) was measured in the midsagittal plane by drawing 2 lines, one over the frontal soft tissue and another over the nasal soft tissue, bringing the angle right over the nasion (Fig. 3). Inter-frontozygomatic suture (FZS) distance was measured in the 3-dimensional image facing forward. The suture was identified on the right and left sides, and measurements were made from the orbital end of the suture lines. Statistical Analysis. Statistical analysis was performed using Statistical Package for the Social Sciences software (SPSS) and Stata 12.1 (Stata Corp., College Station, TX, U.S.A.). Paired samples t tests were used to determine differences in traits within individuals while analysis of variance (ANOVA) was used to determine differences between ethnic groups. Linear regression models were used for univariate and multivariate analysis. A p value of less than 0.05 was considered as statistically significant.
RESULTS The patient demographics by age and race descriptions are shown in Table 1. Two main racial groups were analyzed: “group 1” and “group 2.” Patients in “group 1” included races that have mongoloid features namely Southeast Asians (SEA), including Chinese, Vietnamese, Malay, Indonesian, and “group 2” consisted of races that have Caucasianoid features namely South Asians (Indian, Bangladeshi, and Pakistani) and Occidental races. The 2 groups were chosen based on the facial features. SEA generally have a rounded and broader face and a less pronounced,
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
323
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
A. Mei-Yen Yong et al.
FIG. 2. Sagittal section showing the nasolacrimal duct measurement. broad nasal bridge, whereas South Asians and Occidentals tend to have a narrower face and a more defined, narrow nasal bridge (Table 1). The mean (standard deviation) age for group 1 was 42.75 (16.79) and for group 2 was 36.24 (15.58). The male/female ratio for group 1 was 34/31, and for group 2, it was 10/15. There was no statistically significant difference for age (p = 0.09) and for gender proportion (p = 0.295) between 2 groups. The authors analyzed the inter- and intraobserver correlation of the data collected by 2 independent observers to prove its reliability and reproducibility. This is shown in Table 2, suggesting substantial to excellent correlation (0.707–0.917) across all measurements. The vertical lacrimal fossa diameter, ALC thickness, and narrowest diameter of the NLD did not differ significantly between the right and left sides, and the overall mean diameters are shown in Table 3. The ANOVA test showed that the inter-FZS distance varied significantly among ethnic groups (p = 0.025; Table 3). However, there were no significant differences in all other measurements studied. In univariate analysis, the ALC thickness was inversely associated with FNA (p = 0.034). This was further confirmed in multivariate analysis, whereby a smaller FNA was consistent with a thicker ALC (p = 0.026). For every decrease in FNA of 1°, the ALC thickness increased by 0.024 (Table 4). Similarly, the ALC and inter-FZS distance showed a significant correlation in univariate analysis (p = 0.034) and in multivariate analysis (p = 0.026), indicating that a broader face in SEA population is more likely to have thicker ALC (Table 4). There was a significant association of ALC thickness with narrowest NLD diameter in univariate analysis (p = 0.006), which was confirmed in multivariate analysis (p = 0.019). It showed that the wider the NLD diameter, the thicker the ALC thickness (Table 4).
324
In addition, a wider NLD diameter was highly consistent with a greater inter-FZS distance (p = 0.001), suggesting that group 2 with narrower faces had narrower NLD (Table 5).
DISCUSSION For sufficient osteotomy in DCR, manipulation of intranasal structures to make a good ostium might cause complications including bleeding and scarring of the nasal mucosa. Comprehensive understanding and prediction of structural variation are critical to surgical success. This emphasizes the importance of awareness of lacrimal bone characteristics in patients of different ethnic groups. In this series, racial differences were seen in the inter-FZS distance among SEA, South Asians, and Caucasians. This proves the general consensus that the narrower face represents Caucasianoid features. The NLD diameter is not only narrower in women and in patients with NLDO15,16 but also significantly different in some races. McCormick and Sloan17 showed no differences in narrowest NLD diameter between Caucasians and New Zealand Maoris (3.7 mm), but the narrowest NLD was significantly wider in the Pacific race (4.1 mm) compared with the above groups. They used axial CT scans to make measurements. There are no studies which compare Asians with Caucasians. This study shows that there is no racial variation in the groups that the authors compared, and the diameter was similar to the former group in the above study. NLD diameter was significantly narrower in people with narrower faces. The fact that the patients with Caucasianoid features had significantly narrower faces may mean that this
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Bony Nasolacrimal Parameters in Asians
FIG. 3. Sagittal CT image at midsagittal plane showing frontonasal angle measurement, as verified in 3 planes.
TABLE 1. Patient demographics by gender and ethnicity description Age Mean (SD) Median
Gender
Ethnicity
40.9 (15.72)
Men
44 (48.9%)
Group 1
42
Women
46 (51.1%)
Group 2
Southeast Asian (Chinese, Malay, Indonesian, Vietnamese) South Asian (Indian, Bangladeshi, Pakistani), Caucasian, Occidental races
65 (72.2%) 25 (27.8%)
SD, standard deviation.
group is more predisposed to NLD obstruction. A clinical study of the incidence of NLDO in different races will be necessary to prove this theory. It is not just the diameter but the angle of bony nasolacrimal canal may also play an important role in the occurrence of NLD obstruction. This is evident in a study by Takahashi et al.18 who found that the horizontal angle of NLD differs among individuals with equal split between medial and lateral inclinations about sagittal plane, implying that a larger bending lacrimal passage may cause tear flow stagnation resulting in NLD obstruction. Similarly, although the ALC appeared clinically thicker in SEA than South Asians and Caucasians during DCR, this study showed no racial differences in ALC thickness. Lui19 describes 2 studies, one measuring thickness of ALC involving 50 Chinese skulls and another measuring the thickness of lacrimal bone during DCR procedure in Chinese population. The average thickness reported is 4.65 mm in the former study and 5.8 ± 0.9 mm in men and 4.2 ± 0.8 mm in women in the latter
study. Because the methodology is not described, it is not clear which part of the bony area was measured in the above studies. Nevertheless, these figures are in stark contrast to the study by Hartikainen et al.,20 who measured the lacrimal bone thickness in Western population and found that the average thickness was 1.06 mm. The figures in the current study show a higher average value (5.77–6.13 mm) because the measurements were aimed at ALC and not lacrimal bone. Lacrimal bone thickness does not have a direct impact on the external and endoscopic DCR surgery, but it would be relevant for a transcanalicular laser DCR surgery. Univariate and multivariate analysis of these data revealed that smaller FNA was associated with increased ALC thickness. Irrespective of the race, a more acute FNA may predict increased difficulty in osteotomy, requiring additional instrumentation. Thus, although the broad nasal bridge of SEA patients makes an incision more visible from a frontal view than in a
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
325
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
A. Mei-Yen Yong et al.
TABLE 2. Interobserver correlation Measurements Lacrimal fossa vertical diameter (mm) Right Left Narrowest NLD diameter (mm) Right Left Anterior lacrimal crest (mm) Right Left Frontonasal angle (°) Inter-FZS distance (mm)
ICC
95% CI
0.706 0.794
0.430–0.863 0.564–0.909
0.901 0.791
0.784–0.957 0.535–0.910
0.875 0.737 0.825 0.917
0.692–0.937 0.445–0.884 0.630–0.922 0.746–0.968
CI, confidence interval; FZS distance, frontozygomatic suture distance; ICC, intraclass coefficient; NLD, nasolacrimal duct.
South Asian or Caucasian patient, it is contrary to the belief that a larger ALC bone thickness is observed in patients with a lower nasal bridge, whom the authors assumed would have a
more obtuse FNA. It is possible that the authors are using the wrong parameter to look at the racial differences in ALC thickness because FNA measurement is dependent not only on nasal projection but also on frontal projection. Narioka et al.21 made an interesting observation that the nasal depth can be used to predict the angle between the lacrimal sac and NLD (LS-NLD) in the sagittal plane, and nasal height and length can be used to predict the LS-NLD angle in the coronal plane. Future studies may be designed to measure the bony nasal projection, nasofacial angle, and intercanthal distance, which may better predict ALC thickness. This study is limited primarily by the smaller number of patients in the non-SEA group, and the division of subjects into 2 groups is based on certain assumptions. No distinction is made between different sexes and different age groups, which may affect the parameters measured in this study. Further comparative studies are needed to fully address the anatomic differences, which would provide better surgical guidelines. The key to being a successful lacrimal surgeon lies in the awareness and ability to adjust surgical technique when anatomic variation is encountered.
TABLE 3. Results of statistical analysis for differences in measurement among various ethnic groups ANOVA test for difference between ethnic groups Group 1, mean (SD)
Group 2, mean (SD)
F test, p value
17.15 (2.75) 17.03 (2.79)
17.12 (2.64) 17.27 (2.53)
0.961 0.751
3.82 (0.90) 3.78 (0.75)
3.54 (1.02) 3.81 (1.11)
0.254 0.889
5.78 (2.39) 6.13 (2.15) 133.78 (10.23) 103.08 (4.96)
0.950 0.438 0.203 0.025
Lacrimal fossa diameter (mm) OD OS Narrowest NLD diameter (mm) OD OS Anterior lacrimal crest (mm) OD OS Frontonasal angle (°) Inter-FZS distance (mm)
5.81 (1.84) 5.77 (1.61) 136.82 (8.50) 106.16 (5.08)
Group 1: Southeast Asians; Group 2: South Asians and Caucasians. FZS distance, Frontozygomatic suture distance; NLD, nasolacrimal duct; SD, standard deviation.
TABLE 4. Association with ALC thickness Univariate Measurements Inter-FZS distance (mm) Frontonasal angle (°) Lacrimal fossa diameter (mm) Narrowest NLD diameter (mm)
Multivariate*
β
95% CI
p value
β
95% CI
p value
0.035 −0.024 −0.031 0.328
−0.004 to 0.073 −0.045 to −0.002 −0.103 to 0.041 0.098 to 0.557
0.034 0.034 0.393 0.006
0.007 −0.024 — 0.312
−0.037 to 0.051 −0.046 to −0.003 — 0.053 to 0.572
0.026 0.026 — 0.019
*Adjusted for ethnic groups. ALC, anterior lacrimal crest; CI, confidence interval; FZS distance, frontozygomatic suture distance; NLD, nasolacrimal duct.
TABLE 5. Association with frontonasal angle and inter-FZS distance Frontonasal angle (°)
Model 1* Anterior lacrimal crest (mm) Model 2* Narrowest NLD diameter (mm)
Inter-FZS distance (mm)
β
95% CI
p value
β
95% CI
p value
−2.18
−4.18 to −0.18
0.033
1.03
−0.09 to 2.15
0.072
0.16
−2.17 to 2.48
0.895
2.86
1.74 to 3.97
Assessment of Bony Nasolacrimal Parameters Among Asians Adeline Mei-Yen Yong*, Daniel Binhang Zhao, M.Sc†, Shuen Chao Siew, F.R.C.S.‡, Poh Sun Goh, F.R.C.R.†, Jiemin Liao, M.A.§║, and Shantha Amrith, F.R.C.S.§ *Yong Loo Lin School of Medicine, National University of Singapore, Singapore; †Department of Diagnostic Imaging, ‡Department of Otolaryngology, and §Department of Ophthalmology, National University Hospital, Singapore; and ║Singapore Eye Research Institute, Singapore
Purpose: A case series evaluating racial differences in the nasolacrimal region and quantifying the anterior lacrimal crest thickness and minimum nasolacrimal duct diameter in Asians. Methods: Facial or orbital CT scans of 90 consecutive patients were retrospectively reviewed. Evidence of lacrimal fossa tumor or trauma excluded a patient. Using 3-dimensional image software, the thickness of the anterior lacrimal crest, narrowest diameter of the nasolacrimal duct, vertical diameter of the lacrimal sac fossa, frontonasal angle, and inter-frontozygomatic suture distance were measured in axial, sagittal, and coronal planes. Results: Inter- and intraobserver correlation of a sample data proved reliability via intraclass correlation coefficient (0.706– 0.917). Southeast Asians had a wider inter-frontozygomatic suture distance than South Asian and Occidental races (p = 0.025). Vertical lacrimal fossa diameter, anterior lacrimal crest thickness, and narrowest nasolacrimal duct diameter did not differ significantly between right and left sides or among ethnic groups. Narrower nasolacrimal duct diameter was significantly associated with decreased inter-frontozygomatic suture distance (p < 0.001), namely in patients with narrower faces. The anterior lacrimal crest thickness was significantly affected by the nasal configuration and thicker in patients with more acute frontonasal angle (p = 0.026). Conclusions: There is no significant difference in nasolacrimal duct diameter among ethnic groups, which may predispose one to nasolacrimal duct obstruction. But, this is significantly associated with inter-frontozygomatic suture distance, suggesting that a wider face is associated with wider nasolacrimal duct diameter. An acute frontonasal angle predicts a thicker anterior lacrimal crest for surgical preparation during dacryocystorhinostomy. (Ophthal Plast Reconstr Surg 2014;30:322–327)
E
piphora caused by nasolacrimal duct obstruction (NLDO) is a common ophthalmologic problem and accounts for 3% of ophthalmologic clinic visits.1 Obstruction of the nasolacrimal drainage system can be congenital or acquired, which is
Accepted for publication November 26, 2013. The authors have no financial or conflicts of interest to disclose. Address correspondence and reprint requests to Shantha Amrith, F.R.C.S., Department of Ophthalmology, National University Hospital, 5 Lower Kent Ridge Road, Singapore, Singapore 119074. E-mail: [email protected] DOI: 10.1097/IOP.0000000000000101
322
further classified into primary and secondary causes.2 The most frequent causes of secondary acquired NLDO include facial surgery, trauma, neoplasm, sarcoidosis, and Wegener granulomatosis. In contrast, primary acquired NLDO is idiopathic and characterized by gradual chronic inflammation and fibrosis along the NLD, causing increased obstruction. Studies have suggested that racial and sex differences in facial skull dimensions may explain narrower NLDs and hence differences in NLDO incidence.3–6 External dacryocystorhinostomy (DCR) is a standard approach for treating complete NLDO.7–9 Endonasal DCR has advanced with new technology and instrumentation; however, it is still perceived as more challenging than external DCR.10–12 The endonasal approach involves creating an adequately sized and correctly located osteotomy for long-term patency of the nasolacrimal shunt.8,9,13 It is a common belief among oculoplastic surgeons that East Asian bones in the region of anterior lacrimal crest (ALC) are thicker than those of Caucasians.14 Furthermore, performing endonasal DCR on Southeast Asian patients with a low and broad nasal bridge is sometimes more difficult than on South Asian or Caucasian patients with a higher and narrower nasal bridge. Based on such surgical experience, the authors sought to determine the relationship between nose shape and facial anatomy. This is the first study performed to quantify such measurements that may explain the prevalence of NLDO and impact on DCR surgery in Asians. The authors also attempted to reveal the usefulness of knowing facial anatomy as a predictor of facility in endonasal DCR.
METHODS The authors performed a retrospective review of consecutive facial or orbital CT scans of 90 patients performed at the diagnostic imaging unit at the National University Hospital, Singapore, between January 1, 2010, and December 31, 2011. Evidence of tumor or trauma in the region of the lacrimal fossa excluded a patient. Institutional review board approval for the study was obtained. All scans were collected and numerated; patient identification data were rendered anonymous. CT Technique. All examinations were performed with a 64-slice CT scanner (Brilliance CT; Philips Healthcare, Cleveland, OH, U.S.A.). Contiguous axial 1-mm–thick sections parallel to the infraorbitomeatal line and coronal as well as sagittal scans perpendicular to this plane were acquired with a field of view of 180 mm, matrix size of 512 × 512, and a reconstruction kernel for bone. Measurements. Reconstructed images were loaded in the CT work station (Extended Brilliance Workspace; Philips), and measurements were obtained using calipers on a computer screen, as seen in
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Bony Nasolacrimal Parameters in Asians
FIG. 1. Axial CT image showing anterior lacrimal crest measurement, with correlation across 3 planes.
Figures 1, 2, and 3. The software allows for simultaneous viewing of scans in all 3 planes, and images can be toggled between different planes to verify any chosen point for measurement. Two observers, a medical student (first author) and a fellow in ophthalmic plastic and reconstructive surgery, carried out the measurements independently at different times on the first 23 scans to validate the measurements. Measurement Protocol. All measurements were performed in bone window. Thickness of the ALC was measured in the axial scans. Using the sagittal scan, the upper end of NLD was identified to make sure that the sac region was identified correctly. From that point, the thickest part of the ALC was identified by scrolling up (in axial scan) and anteriorly (in coronal scan; Fig. 1). Measurements were taken in the axial scan after the selected points were repeatedly checked in coronal section. The NLD was identified in the sagittal section, and by scrolling across adjacent cuts, the central axis of the canal was determined and a line was drawn along the long axis. The narrowest diameter was measured perpendicular to the line drawn. The start and the end point of the measure was the inner rim of the canal (Fig. 2). To ensure that the minimum diameter had been measured, several measurements were taken at different levels and the least recorded. Vertical diameter of the lacrimal sac fossa was measured using the 3-dimensional skull image. Reference points at the upper end of the NLD were marked. The point was verified in all 3 types of scans. A line was drawn from this point to where the frontomaxillary suture and the fossa intersected. The measurement of this line gave the vertical diameter of the lacrimal sac.
Frontonasal angle (FNA) was measured in the midsagittal plane by drawing 2 lines, one over the frontal soft tissue and another over the nasal soft tissue, bringing the angle right over the nasion (Fig. 3). Inter-frontozygomatic suture (FZS) distance was measured in the 3-dimensional image facing forward. The suture was identified on the right and left sides, and measurements were made from the orbital end of the suture lines. Statistical Analysis. Statistical analysis was performed using Statistical Package for the Social Sciences software (SPSS) and Stata 12.1 (Stata Corp., College Station, TX, U.S.A.). Paired samples t tests were used to determine differences in traits within individuals while analysis of variance (ANOVA) was used to determine differences between ethnic groups. Linear regression models were used for univariate and multivariate analysis. A p value of less than 0.05 was considered as statistically significant.
RESULTS The patient demographics by age and race descriptions are shown in Table 1. Two main racial groups were analyzed: “group 1” and “group 2.” Patients in “group 1” included races that have mongoloid features namely Southeast Asians (SEA), including Chinese, Vietnamese, Malay, Indonesian, and “group 2” consisted of races that have Caucasianoid features namely South Asians (Indian, Bangladeshi, and Pakistani) and Occidental races. The 2 groups were chosen based on the facial features. SEA generally have a rounded and broader face and a less pronounced,
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
323
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
A. Mei-Yen Yong et al.
FIG. 2. Sagittal section showing the nasolacrimal duct measurement. broad nasal bridge, whereas South Asians and Occidentals tend to have a narrower face and a more defined, narrow nasal bridge (Table 1). The mean (standard deviation) age for group 1 was 42.75 (16.79) and for group 2 was 36.24 (15.58). The male/female ratio for group 1 was 34/31, and for group 2, it was 10/15. There was no statistically significant difference for age (p = 0.09) and for gender proportion (p = 0.295) between 2 groups. The authors analyzed the inter- and intraobserver correlation of the data collected by 2 independent observers to prove its reliability and reproducibility. This is shown in Table 2, suggesting substantial to excellent correlation (0.707–0.917) across all measurements. The vertical lacrimal fossa diameter, ALC thickness, and narrowest diameter of the NLD did not differ significantly between the right and left sides, and the overall mean diameters are shown in Table 3. The ANOVA test showed that the inter-FZS distance varied significantly among ethnic groups (p = 0.025; Table 3). However, there were no significant differences in all other measurements studied. In univariate analysis, the ALC thickness was inversely associated with FNA (p = 0.034). This was further confirmed in multivariate analysis, whereby a smaller FNA was consistent with a thicker ALC (p = 0.026). For every decrease in FNA of 1°, the ALC thickness increased by 0.024 (Table 4). Similarly, the ALC and inter-FZS distance showed a significant correlation in univariate analysis (p = 0.034) and in multivariate analysis (p = 0.026), indicating that a broader face in SEA population is more likely to have thicker ALC (Table 4). There was a significant association of ALC thickness with narrowest NLD diameter in univariate analysis (p = 0.006), which was confirmed in multivariate analysis (p = 0.019). It showed that the wider the NLD diameter, the thicker the ALC thickness (Table 4).
324
In addition, a wider NLD diameter was highly consistent with a greater inter-FZS distance (p = 0.001), suggesting that group 2 with narrower faces had narrower NLD (Table 5).
DISCUSSION For sufficient osteotomy in DCR, manipulation of intranasal structures to make a good ostium might cause complications including bleeding and scarring of the nasal mucosa. Comprehensive understanding and prediction of structural variation are critical to surgical success. This emphasizes the importance of awareness of lacrimal bone characteristics in patients of different ethnic groups. In this series, racial differences were seen in the inter-FZS distance among SEA, South Asians, and Caucasians. This proves the general consensus that the narrower face represents Caucasianoid features. The NLD diameter is not only narrower in women and in patients with NLDO15,16 but also significantly different in some races. McCormick and Sloan17 showed no differences in narrowest NLD diameter between Caucasians and New Zealand Maoris (3.7 mm), but the narrowest NLD was significantly wider in the Pacific race (4.1 mm) compared with the above groups. They used axial CT scans to make measurements. There are no studies which compare Asians with Caucasians. This study shows that there is no racial variation in the groups that the authors compared, and the diameter was similar to the former group in the above study. NLD diameter was significantly narrower in people with narrower faces. The fact that the patients with Caucasianoid features had significantly narrower faces may mean that this
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
Bony Nasolacrimal Parameters in Asians
FIG. 3. Sagittal CT image at midsagittal plane showing frontonasal angle measurement, as verified in 3 planes.
TABLE 1. Patient demographics by gender and ethnicity description Age Mean (SD) Median
Gender
Ethnicity
40.9 (15.72)
Men
44 (48.9%)
Group 1
42
Women
46 (51.1%)
Group 2
Southeast Asian (Chinese, Malay, Indonesian, Vietnamese) South Asian (Indian, Bangladeshi, Pakistani), Caucasian, Occidental races
65 (72.2%) 25 (27.8%)
SD, standard deviation.
group is more predisposed to NLD obstruction. A clinical study of the incidence of NLDO in different races will be necessary to prove this theory. It is not just the diameter but the angle of bony nasolacrimal canal may also play an important role in the occurrence of NLD obstruction. This is evident in a study by Takahashi et al.18 who found that the horizontal angle of NLD differs among individuals with equal split between medial and lateral inclinations about sagittal plane, implying that a larger bending lacrimal passage may cause tear flow stagnation resulting in NLD obstruction. Similarly, although the ALC appeared clinically thicker in SEA than South Asians and Caucasians during DCR, this study showed no racial differences in ALC thickness. Lui19 describes 2 studies, one measuring thickness of ALC involving 50 Chinese skulls and another measuring the thickness of lacrimal bone during DCR procedure in Chinese population. The average thickness reported is 4.65 mm in the former study and 5.8 ± 0.9 mm in men and 4.2 ± 0.8 mm in women in the latter
study. Because the methodology is not described, it is not clear which part of the bony area was measured in the above studies. Nevertheless, these figures are in stark contrast to the study by Hartikainen et al.,20 who measured the lacrimal bone thickness in Western population and found that the average thickness was 1.06 mm. The figures in the current study show a higher average value (5.77–6.13 mm) because the measurements were aimed at ALC and not lacrimal bone. Lacrimal bone thickness does not have a direct impact on the external and endoscopic DCR surgery, but it would be relevant for a transcanalicular laser DCR surgery. Univariate and multivariate analysis of these data revealed that smaller FNA was associated with increased ALC thickness. Irrespective of the race, a more acute FNA may predict increased difficulty in osteotomy, requiring additional instrumentation. Thus, although the broad nasal bridge of SEA patients makes an incision more visible from a frontal view than in a
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
325
Ophthal Plast Reconstr Surg, Vol. 30, No. 4, 2014
A. Mei-Yen Yong et al.
TABLE 2. Interobserver correlation Measurements Lacrimal fossa vertical diameter (mm) Right Left Narrowest NLD diameter (mm) Right Left Anterior lacrimal crest (mm) Right Left Frontonasal angle (°) Inter-FZS distance (mm)
ICC
95% CI
0.706 0.794
0.430–0.863 0.564–0.909
0.901 0.791
0.784–0.957 0.535–0.910
0.875 0.737 0.825 0.917
0.692–0.937 0.445–0.884 0.630–0.922 0.746–0.968
CI, confidence interval; FZS distance, frontozygomatic suture distance; ICC, intraclass coefficient; NLD, nasolacrimal duct.
South Asian or Caucasian patient, it is contrary to the belief that a larger ALC bone thickness is observed in patients with a lower nasal bridge, whom the authors assumed would have a
more obtuse FNA. It is possible that the authors are using the wrong parameter to look at the racial differences in ALC thickness because FNA measurement is dependent not only on nasal projection but also on frontal projection. Narioka et al.21 made an interesting observation that the nasal depth can be used to predict the angle between the lacrimal sac and NLD (LS-NLD) in the sagittal plane, and nasal height and length can be used to predict the LS-NLD angle in the coronal plane. Future studies may be designed to measure the bony nasal projection, nasofacial angle, and intercanthal distance, which may better predict ALC thickness. This study is limited primarily by the smaller number of patients in the non-SEA group, and the division of subjects into 2 groups is based on certain assumptions. No distinction is made between different sexes and different age groups, which may affect the parameters measured in this study. Further comparative studies are needed to fully address the anatomic differences, which would provide better surgical guidelines. The key to being a successful lacrimal surgeon lies in the awareness and ability to adjust surgical technique when anatomic variation is encountered.
TABLE 3. Results of statistical analysis for differences in measurement among various ethnic groups ANOVA test for difference between ethnic groups Group 1, mean (SD)
Group 2, mean (SD)
F test, p value
17.15 (2.75) 17.03 (2.79)
17.12 (2.64) 17.27 (2.53)
0.961 0.751
3.82 (0.90) 3.78 (0.75)
3.54 (1.02) 3.81 (1.11)
0.254 0.889
5.78 (2.39) 6.13 (2.15) 133.78 (10.23) 103.08 (4.96)
0.950 0.438 0.203 0.025
Lacrimal fossa diameter (mm) OD OS Narrowest NLD diameter (mm) OD OS Anterior lacrimal crest (mm) OD OS Frontonasal angle (°) Inter-FZS distance (mm)
5.81 (1.84) 5.77 (1.61) 136.82 (8.50) 106.16 (5.08)
Group 1: Southeast Asians; Group 2: South Asians and Caucasians. FZS distance, Frontozygomatic suture distance; NLD, nasolacrimal duct; SD, standard deviation.
TABLE 4. Association with ALC thickness Univariate Measurements Inter-FZS distance (mm) Frontonasal angle (°) Lacrimal fossa diameter (mm) Narrowest NLD diameter (mm)
Multivariate*
β
95% CI
p value
β
95% CI
p value
0.035 −0.024 −0.031 0.328
−0.004 to 0.073 −0.045 to −0.002 −0.103 to 0.041 0.098 to 0.557
0.034 0.034 0.393 0.006
0.007 −0.024 — 0.312
−0.037 to 0.051 −0.046 to −0.003 — 0.053 to 0.572
0.026 0.026 — 0.019
*Adjusted for ethnic groups. ALC, anterior lacrimal crest; CI, confidence interval; FZS distance, frontozygomatic suture distance; NLD, nasolacrimal duct.
TABLE 5. Association with frontonasal angle and inter-FZS distance Frontonasal angle (°)
Model 1* Anterior lacrimal crest (mm) Model 2* Narrowest NLD diameter (mm)
Inter-FZS distance (mm)
β
95% CI
p value
β
95% CI
p value
−2.18
−4.18 to −0.18
0.033
1.03
−0.09 to 2.15
0.072
0.16
−2.17 to 2.48
0.895
2.86
1.74 to 3.97
