Graefes Arch Clin Exp Ophthalmol DOI 10.1007/s00417-014-2643-1

PEDIATRICS

The prevalence of amblyogenic factors in children with persistent congenital nasolacrimal duct obstruction Bahram Eshraghi & Mohammad Reza Akbari & Masoud Aghsaei Fard & Azadeh Shahsanaei & Raheleh Assari & Arash Mirmohammadsadeghi

Received: 4 January 2014 / Revised: 27 February 2014 / Accepted: 9 April 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Background To evaluate the prevalence of amblyopia risk factors in children that underwent probing for persistent congenital nasolacrimal duct obstruction (CNLDO). Methods The medical records of children with CNLDO (after 1 year of age) that underwent probing were reviewed. Amblyopia risk factors, based on the American Association for Pediatric Ophthalmology and Strabismus referral criteria in 2013, were sought in the patients records before probing. The proportion of the patients with anisometropia >1.5 diopters (D) was separately calculated. In unilateral cases of CNLDO, sphere, astigmatism, and spherical equivalent of the eyes with CNLDO were compared with contralateral eyes in order to assess the effect of CNLDO on refractive error. In the followup examinations, the success of the probing or the need for additional procedures (Crawford intubation, Monoka intubation, or dacryocystorhinostomy) was evaluated. The prevalence of anisometropia between ‘successful probing’ and ‘failed probing’ groups was compared. Results A total of 433 cases were included in the study. 41 cases (9.46 %) had amblyopia risk factors. Twenty-four cases (5.5 %) had anisometropia >1.5 D (spherical or cylindrical). In unilateral cases of CNLDO, the sphere and spherical equivalent of the eyes with CNLDO were significantly greater than those of the contralateral eyes (p1.5 D (compared with the prevalence in the general B. Eshraghi : M. R. Akbari : M. A. Fard : A. Shahsanaei : R. Assari : A. Mirmohammadsadeghi (*) Farabi Eye Research Center, Tehran University of Medical Sciences, Tehran, Iran e-mail: [email protected]

population) and significantly greater sphere and spherical equivalents in the eye with CNLDO (compared with contralateral eye) in unilateral cases with CNLDO, suggested some relationships between anisometropia and long-term untreated CNLDO. The finding of more anisometropia in failed probing cases may support the theory of structural abnormality as an explanation for the possible relationship between congenital nasolacrimal duct obstruction and anisometropia. Keywords Congenital nasolacrimal duct obstruction . Anisometropia . Amblyopia risk factors

Introduction About 20–30 % of all newborns have symptoms of congenital nasolacrimal duct obstruction (CNLDO) [1]. Ninety-six percent of CNLDOs spontaneously resolve in the first year of life [2]. Approximately 4 % of patients have persistent symptoms after 1 year of age [2]. The standard surgical procedure for persistent CNLDO is lacrimal system probing [1]. Several studies have been done to evaluate the relationship between CNLDO and amblyopia risk factors (including anisometropia) [3–12]. Most of these studies were done on the cases with CNLDO from birth and did not focus on persistent cases of CNLDO that required probing [3–12]. So, the present study was planned to evaluate the prevalence of amblyopia risk factors in persistent CNLDO cases that underwent probing (Fig. 1).

Methods The study was done at the Farabi Eye Hospital between 2009 and 2013. Retrospectively, the medical records of children

Graefes Arch Clin Exp Ophthalmol

with CNLDO that underwent probing were reviewed. Probing in our center had been done only after documenting persistent symptoms after 1 year of age or the occurrence of acute dacryocystitis at a younger age. The records that did not have complete information (including cycloplegic refraction) were excluded from the study. The study and data collection were compliant with the principles of the Declaration of Helsinki and conformed to all local laws. Age at the time of the probing, sex and affected eye were noted in the records. The data of cycloplegic refraction before probing was evaluated. Cycloplegic refraction was done (with cyclopentolate 1 %) by an optometrist that was highly skilled in retinoscopy of the children. Other ophthalmologic examinations were done for all patients by one oculoplastic specialist (B.E.). The diagnosis of manifest strabismus was done by cover test at far and near. Followup examinations were also reviewed and the success of the probing or the need for additional procedures (Crawford intubation, Monoka intubation, or dacryocystorhinostomy) was evaluated. The diagnosis of CNLDO was done with a history of tearing and/or discharge, reflux of tear with pressure on lacrimal sac and a positive dye disappearance test. Differential diagnosis such as lid abnormalities (trichiasis, entropion,…), ocular allergy and congenital glaucoma were ruled out in all patients by complete eye examinations and, if needed, examination under anesthesia. Amblyopia risk factors, based on the American Association for Pediatric Ophthalmology and Strabismus (AAPOS) referral criteria in 2013 [13], were sought in the patients records before probing. These risk factors are summarized in Table 1. The proportion of the patients with anisometropia >1.5 diopters (D) was separately calculated to compare the results with other studies [14]. Furthermore, in unilateral cases of CNLDO, sphere, astigmatism, and spherical equivalent of eyes with and without CNLDO were compared to assess the effect of CNLDO on refractive error. The patients were divided into two subgroups for further analysis: “successful probing and failed probing”. “Successful probing” group was defined as cases with no tearing and discharge after probing. The “failed probing” group was defined as cases with persistent tearing or discharge after probing that required other interventions. For better analysis, three variables were created to evaluate degree of anisometropia in unilateral cases of CNLDO: diff-sphere, diff-astig, and diff-sphereq. These variables demonstrated the amount of difference between hyperopia, astigmatism, and spherical equivalent of the eye with CNLDO and the eye without CNLDO, respectively. They were calculated from the following formulas:

Diff sphere ¼ spherical part of refractive error in eye with CNLDO spherical part of refractive error in eye without CNLDO Diff astig ¼ astigmatic part of refractive error in eye with CNLDO astigmatic part of refractive error in eye without CNLDO Diff sphereq ¼ spherical equivalent in eye with CNLDO spherical equivalent in eye without CNLDO

The goal of creating these variables was to compare the degree of anisometropia in ‘successful probing’ and ‘failed probing’ groups. So, if both eyes had similar signs (both plus or both minus), only absolute values (without plus or minus signs) were used in the calculations. Statistical analysis was done by SPSS version 20 (SPSS inc., Chicago, IL). To evaluate the difference of anisometropia prevalence between subgroups, the chi-square test was used. To compare refractive errors between the eyes with and without CNLDO, the paired t-test was performed. To evaluate differences of computed variables between subgroups, the two-sample t-test was performed. A P1.5 D. In all unilateral probing cases with anisometropia (both definitions), the same eye that underwent probing, also had higher hyperopia or astigmatism. In unilateral cases of CNLDO, the sphere and spherical equivalent of the eyes with CNLDO were significantly greater than those of the contralateral eyes (p1.5 D) prevalence in the failed probing group (compared with successful probing group) at the initial examination (p=0.03). The mean age of two groups did not show a significant difference (p=0.50). The proportion of successful and failed probing cases in three refractive amblyopia risk factor groups were shown in Fig. 1 In unilateral cases of CNLDO, three computed variables (diff-sphere, diff-astig, and diff-sphereq) were analyzed to compare the degree of anisometropia between the two groups. The mean and standard deviation of these computed variables in successful probing and failed probing groups were shown in Table 3. Both diff-sphere and diffsphereq were significantly greater in the failed probing group (compared with successful probing group) (p value=0.02 and 0.04, respectively). But there was no significant difference in diff-astig between the two groups (p value=0.8).

Discussion In the past centuries, CNLDO had been considered a benign disorder, with no effect on visual development. In the last two decades, several studies were done to evaluate the possible Table 1 Amblyopia risk factors, based on the American Association for Pediatric Ophthalmology and Strabismus (AAPOS) original referral criteria in 2013 (D = diopters, PD = prism diopters, PP = primary position) Refractive risk factors

Non-refractive risk factors

Age of children

All ages

Amblyopia risk factors

12–30 months astigmatism >2.0 D Manifest strabismus >8 PD in PP hyperopia >4.5 D Media opacity >1 mm anisometropia >2.5 D myopia >−3.5 D 31–48 months astigmatism >2.0 D hyperopia >4.0 D anisometropia >2.0 D myopia >−3.0 D >48 months astigmatism >1.5 D hyperopia >3.5 D anisometropia >1.5 D myopia >−1.5 D

Graefes Arch Clin Exp Ophthalmol Table 2 Types of amblyopia risk factors (based on 2013 AAPOS referral criteria) and number of the patients in each group Amblyopia risk factor type

Number of patients

Anisometropic hyperopia Bilateral high hyperopia Anisometropic astigmatism Bilateral high astigmatism

16 10 6 5

Strabismus Total

4 41

relationship between CNLDO and amblyopia risk factors [3–12]. Our study had some similarities with other studies, but evaluated amblyogenic factors in a different ethnicity and added some new concepts such as probing failure and computed variables. In the study of Piatrowski and colleagues on patients with persistent symptoms of CNLDO, the prevalence of anisometropia (>1 D) was 9.8 % [3]. In contrast with our study, only 66.2 % of their patients underwent surgical procedures for symptoms of CNLDO [3]. In the study of Matta and colleagues on patients with CNLDO, using guidelines of the original (2003) AAPOS referral criteria [13], the prevalence of amblyopia risk factors was 22 % and the prevalence of anisometropia > 1.5 D was 2.4 % [4]. In another article by Matta and Silbert, they stated that 63 % of previous cases developed amblyopia [5]. In another study by Kipp and colleagues on cases of CNLDO, at initial examination 6.5 % had anisometropia (>1 D) that increased to 8.6 % on subsequent examinations [6]. Other studies by Bagheri and colleagues [7], Simon and colleagues [8], Chalmers and Griffiths [9], Lacey and colleagues [10] and Kim and colleagues [11] were done with different inclusion criteria, different sample sizes, and different definitions of anisometropia or amblyopia [7–11]. On the other hand, Ellis and

Table 3 Mean, standard deviation and p-value of comparing computed variables (diff-sphere and diff-sphereq) in successful probing and failed probing groups in unilateral NLDO cases (SD = standard deviation, Diffsphere = spherical part of refractive error in eye with NLDO- spherical part of refractive error in eye without NLDO, Diff-sphereq = spherical equivalent in eye with NLDO- spherical equivalent in eye without NLDO, Diff-astig = astigmatic part of refractive error in eye with CNLDO- astigmatic part of refractive error in eye without CNLDO) Successful probing Failed probing p-value Diff-sphere +0.089±0.474 (mean ± SD) (diopter) Diff-sphereq +0.092±0.448 (mean ± SD) (diopter) Diff-astig −0.015±0.239 (mean ± SD) (diopter)

+0.343±0.871 0.02 +0.296±0.790 0.04 +0.010±0.513 0.80

colleagues found no relationship between NLDO and amblyopia [12]. In our study, all the cases were 1 year of age or older, had no resolution of CNLDO symptoms and underwent probing. So, the average age of our cases (20.3 months) was higher than other studies (except the study of Bagheri and colleagues) [3–12]. Like other studies, the number of male and female cases was approximately equal [3–12]. Laterality of our cases was the same as other studies (approximately 1/3 right eye, 1/3 left eye, and 1/3 bilateral) [4, 15, 16]. All mentioned studies (except the study of Ellis and colleagues) demonstrated that in their series of CNLDO cases, anisometropia was more prevalent than the general population [3–12]. The differences in the rate of anisometropia between different studies may be the result of different definitions (>1 D or >1.5 D) [3–12]. The mean prevalence of anisometropia (greater than 1 D) in the general population at various ages is approximately 2 % [17]. The prevalence of anisometropia (>1.5 D) in our study was 5.5 %. If we had included the cases with anisometropia between 1 and 1.5 D, the prevalence would have increased. Therefore, our study prevalence was more than the prevalence in the general population (2 %) and Matta and colleagues study (2.4 %). This effect may be due to relationship of persistent CNLDO with anisometropia or excluding cases under 1 year of age. Anisometropia prevalence may increase with increasing age of the patients [17–22]. This idea was demonstrated in the studies of Atkinson and Braddick [17–19], Weale [17, 20], Bourne and colleagues [17, 21] and Quek and colleagues [17, 22]. On the other hand, in our series, in all unilateral probing cases with anisometropia, the same eye that underwent probing, also had higher hyperopia or astigmatism. In addition, in unilateral cases with CNLDO, sphere and spherical equivalent of the eyes with CNLDO were significantly greater than the opposite eyes. These findings suggested some relationship between anisometropia and persistent CNLDO. Therefore, in our study, among amblyopia risk factors, anisometropia may be related to longterm untreated CNLDO. In some studies, the prevalence of total amblyopia risk factors in childhood was approximately 15–20 % [13], but the prevalence of amblyopia was about 2 % [13]. Thus, the majority of the children with amblyopia risk factors never develop amblyopia [13]. The prevalence of total amblyopia risk factors in our study (9.46 %) was less than the general pediatric population [13]. This effect may be due to different characteristics of our ethnicity, excluding cases under 1 year of age, higher mean age of our patients, and previous resolution of amblyopia risk factors in some of them. Matta and Silbert stated that in the follow-up of CNLDO cases, amblyopia risk factors resolved spontaneously in the follow-up of 11–135 months

Graefes Arch Clin Exp Ophthalmol

(mean: 29 months) in 37 % of cases [5]. In their study, strabismus resolved in two children, astigmatism resolved in 14 children and hyperopia resolved in eight children [5]. Moreover, in the Congenital Esotropia Observational Study, among 170 patients of congenital esotropia, 46 (27 %) spontaneously resolved to within 8 PD of orthotropia either with or without spectacle correction [23]. Anyway, in spite of finding some possible relationships between anisometropia and CNLDO, we could not find any relationship between total amblyopia risk factors and persistent CNLDO in children 1 year of age or older. Increased prevalence of anisometropia cases (>1.5 D) in the failed probing group (compared with the successful probing group) was the other interesting finding in our study. This finding was not related to the ages of children because the ages of two groups had no significant differences. Also, the significantly greater amount of diff-sphere and diff-sphereq in the failed probing group means significantly more hyperopic eyes with CNLDO in this group, compared with successful probing group (Table 3). As we know, there were three main theories to explain the relationship between CNLDO and anisometropia [6, 9]: 1Image deformation (due to excessive tear film) may cause inappropriate emmetropization in the involved eye [6]; 2Associated structural abnormality, like maldevelopment of one side of the face, may cause both significant refractive error and abnormal nasolacrimal duct [6, 9]; and 3- Inaccurate refractive error measurements in the presence of excessive tear film [6]. The first and third theories cannot explain the difference of anisometropia (prevalence and degree) between failed probing and successful probing groups. Image deformation and inaccurate measurements can occur in both groups. On the other hand, in our study, probing failures were due to a complex obstruction of the nasolacrimal duct or canalicular stenosis. This complex obstruction or canalicular stenosis may be a part of a larger structural abnormality that causes significant refractive error on the same side. So, the finding of more anisometropia cases in the “failed probing” group may support the second theory as the explanation of a possible relationship between CNLDO and anisometropia. The limitations of our study were its retrospective nature, the absence of long-term refractive follow-up data (after resolution of CNLDO) in the records, and excluding infants under 1 year of age. In spite of these limitations, the relatively large sample size was a benefit for our study to find significant differences between variables. In conclusion, in our series of CNLDO cases that required probing, the prevalence of anisometropia (>1.5 D) was more than the general population and some other studies. This finding and the finding of significantly greater sphere and spherical equivalents in the eyes with CNLDO (compared to contralateral eyes) in unilateral

cases with CNLDO, suggested some relationship between anisometropia and long-term untreated CNLDO. Earlier probing might decrease the formation of anisometropia. The finding of more anisometropia in failed probing cases may support the theory of structural abnormality as an explanation for the possible relationship between CNLDO and anisometropia. So, the clinicians must pay particular attention to cycloplegic refraction of cases with CNLDO, especially failed probing patients, in order to avoid anisometropic amblyopia.

Conflict of interest The authors report no conflict of interest.

References 1. Maheshwari R (2005) Results of probing for congenital nasolacrimal duct obstruction in children older than 13 months of age. Indian J Ophthalmol 53:49–51 2. MacEwen CJ, Young JDH (1991) Epiphora during the first year of life. Eye 5:596–600 3. Piotrowsky JT, Diehl NN, Mohney BG (2010) Neonatal dacryostenosis as a risk factor for anisometropia. Arch Ophthalmol 128:1166–1169 4. Matta NS, Singman EL, Silbert DI (2010) Prevalence of amblyopia risk factors in congenital nasolacrimal duct obstruction. J AAPOS 14: 386–388 5. Matta NS, Silbert DI (2011) High prevalence of amblyopia risk factors in preverbal children with nasolacrimal duct obstruction. J AAPOS 15:350–352 6. Kipp MA, Kipp MA Jr, Struthers W (2013) Anisometropia and amblyopia in nasolacrimal duct obstruction. J AAPOS 17:235–238 7. Bagheri A, Safapoor S, Yazdani S, Yaseri M (2012) Refractive state in children with unilateral congenital nasolacrimal duct obstruction. J Ophthalmic Vis Res 7:310–315 8. Simon JW, Ngo Y, Ahn E, Khachikian S (2009) Anisometropic amblyopia and nasolacrimal duct obstruction. J Pediatr Ophthalmol Strabismus 46:182–183 9. Chalmers RJ, Griffiths PG (1996) Is congenital nasolacrimal duct obstruction a risk factor for the development of amblyopia? Br Orthopt J 53:29–30 10. Lacey BA, McGinnity GF, Johnston PB, Archer DB (1995) Congenital epiphora as a potential cause of amblyopia. Vis Res 35:130 11. Kim JW, Lee H, Chang M, Park M, Lee TS, Baek S (2013) Amblyopia risk factors in infants with congenital nasolacrimal duct obstruction. J Craniofac Surg 24:1123–1125 12. Ellis JD, MacEwen CJ, Young JD (1998) Can congenital nasolacrimal-duct obstruction interfere with visual development? A cohort case control study. J Pediatr Ophthalmol Strabismus 35: 81–85 13. Donahue SP, Arthur B, Neely DE et al (2013) Guidelines for automated preschool vision screening: a 10-year, evidence-based update. J AAPOS 17:4–8 14. Donahue SP, Arnold RW, Ruben JB (2003) Preschool vision screening: what should we be detecting and how should we report it? Uniform guidelines for reporting results of preschool vision screening studies. J AAPOS 7:314–316 15. Thongthong K, Singha P, Liabsuetrakul T (2009) Success of probing for congenital nasolacrimal duct obstruction in children under 10 years of age. J Med Assoc Thail 92:1646–1650

Graefes Arch Clin Exp Ophthalmol 16. Castelo Branco Neto E, Castelo Branco B, Cardoso CC et al (2009) Management of congenital nasolacrimal duct obstruction. Arg Bras Oftalmol 72:75–78 17. Donahue SP (2006) Relationship between anisometropia, patient age, and the development of amblyopia. Am J Ophthalmol 142:132–140 18. Atkinson J, Braddick O (1983) Vision screening and photorefraction—the relation of refractive errors to strabismus and amblyopia. Behav Brain Res 10:71–80 19. Atkinson J, Braddick O, Durden K et al (1984) Screening for refractive errors in 6–9 month old infants by photorefraction. Br J Ophthalmol 68:105–112

20. Weale RA (2002) On the age-related prevalence of anisometropia. Ophthalmic Res 34:389–392 21. Bourne RR, Dineen BP, Ali SM et al (2004) Prevalence of refractive error in Bangladeshi adults: results of the National Blindness and Low Vision Survey of Bangladesh. Ophthalmology 111:1150–1160 22. Quek TP, Chua CG, Chong CS et al (2004) Prevalence of refractive errors in teenage high school students in Singapore. Ophthalmic Physiol Opt 24:47–55 23. Pediatric Eye Disease Investigator Group (2002) Spontaneous resolution of early-onset esotropia: experience of the Congenital Esotropia Observational Study. Am J Ophthalmol 133:109–118

The prevalence of amblyogenic factors in children with persistent congenital nasolacrimal duct obstruction.

To evaluate the prevalence of amblyopia risk factors in children that underwent probing for persistent congenital nasolacrimal duct obstruction (CNLDO...
205KB Sizes 0 Downloads 3 Views