Association Between Preoperative Ocular Parameters and Myopic Shift in Children Undergoing Primary Intraocular Lens Implantation

Abstract

Purpose: To evaluate the association between preoperative ocular parameters and myopic shift following primary intraocular lens (IOL) implantation in pediatric cataracts.

Methods: Eyes from pediatric patients undergoing bilateral cataract surgery with primary IOL implantation were included. Eyes were grouped by age at surgery and subdivided into three axial length (AL) subgroups and three keratometry subgroups. Mixed-effects linear regression was utilized to assess the trend in myopic shift among subgroups. Multivariable analysis was performed to determine factors associated with myopic shift.

Results: A total of 222 eyes were included. The median age at surgery was 4.36 years (interquartile range [IQR], 3.16–6.00 years) and the median follow-up was 4.18 years (IQR, 3.48–4.64 years). As preoperative AL increased, a decreased trend was observed in myopic shift and rate of myopic shift (P = 0.008 and P = 0.003, respectively, in the 4 to <6 years old group; P = 0.002 and P < 0.001, respectively, in the ≥6 years old group). Greater myopic shift and rate of myopic shift were associated with younger age at surgery (P = 0.008 and P = 0.008, respectively). Both myopic shift and rate of myopic shift were negatively associated with AL.

Conclusions: Age at surgery and preoperative AL were associated with myopic shift in pediatric cataracts following primary IOL implantation. Adjusting the target refraction based on preoperative AL could potentially improve patients’ long-term refractive outcome.

Translational Relevance: This study may help to guide the selection of postoperative target refraction according to age at surgery and preoperative ocular parameters for pediatric cataracts.

Introduction

Intraocular lens (IOL) implantation is a common visual rehabilitation method for pediatric cataracts.¹ Growth of children's eyeballs following IOL implantation can result in myopic shift over time which exhibits significant interindividual variability.^2,3 To prevent high myopia in adulthood, children are usually left moderately hyperopic at the time of IOL implantation.⁴ Nevertheless, due to the variability of myopic shift, choosing the optimum postoperative target refraction and IOL power for pediatric patients is still challenging. 

Numerous reports have shown that children receiving IOL implantation at a younger age tend to experience a greater myopic shift.^5–8 Therefore, the determination of postoperative target refraction for children is predominantly based on their age at surgery. In cases of unilateral cataracts, it is also essential to consider the refractive status of the healthy eyes to avoid postoperative anisometropia.^9,10 However, preoperative ocular parameters, including axial length (AL) and cornea curvature in pediatric cataracts, are variable.^11,12 This interindividual variability has not been taken into consideration when determining target refraction based only on the child's age. The effect of preoperative ocular parameters on myopic shift in children following IOL implantation remains unclear. Taking AL as an example, the range of preoperative AL of pediatric cataracts was large even in individuals of the same age.¹¹ Some reports revealed that a shorter AL was related to greater myopic shift,^13–15 and another study demonstrated that, rather than AL, a flatter cornea was associated with greater myopic shift.¹⁶ 

Thus, we conducted this study to further investigate the association between preoperative ocular parameters and myopic shift following IOL implantation in pediatric cataracts. By identifying that, it is possible to adjust the target refraction for children based on their ocular parameters, which could potentially improve the long-term refractive outcome for them. 

Methods

This study was approved by the Zhongshan Ophthalmic Center (Guangzhou, China) Institutional Review Board (2013PRLL001) and conformed to the tenets of Declaration of Helsinki. 

Participants and Study Design

A retrospective review was conducted on children who were diagnosed with cataract and received surgical treatment at Zhongshan Ophthalmic Center between 2010 and 2019. Children were eligible for inclusion if (1) they were diagnosed as congenital or developmental cataract and underwent bilateral simultaneous cataract surgery with primary IOL implantation, and (2) they had follow-up of ≥3 years from the time of surgery. Exclusion criteria were (1) being diagnosed as traumatic cataract or having a history of ocular trauma; (2) having undergone unilateral or secondary IOL implantation; and (3) having pre-existing ocular comorbidities, including but not restricted to, corneal opacity, glaucoma, persistent fetal vasculature, fundus diseases, and so on. 

Both eyes of each subject were included in the analysis. To control the effect of age on the myopic shift, eligible eyes were divided into three groups according to age at surgery: 2 to <4 years old, 4 to <6 years old, and ≥6 years old. Subsequently, eyes within each age group were subdivided into three AL subgroups based on the middle percentiles of the preoperative AL distribution (AL below the 25th percentile, AL between the 25th and 75th percentiles, and AL above the 75th percentile). Using the same methodology, eyes included within each age group were also divided into three average keratometry (AK) subgroups. The trends in myopic shift and in the rate of myopic shift among the different subgroups within each age group were assessed. 

Surgical Techniques

All of the surgeries were performed using standardized procedures, including the creation of a 3.2-mm scleral tunnel or corneal incision at 12 o'clock, anterior capsulorhexis, hydrodissection, lens irrigation/aspiration, primary posterior capsulectomy, and anterior vitrectomy. Commonly, IOLs were implanted in the capsular bag; otherwise, IOLs were implanted in the ciliary sulcus. All of the surgeries were performed under general anesthesia. Initial postoperative target refractions for children were as follows: +2 to +3 D in children 2 to <4 years old, +1 to +1.5 D in children 4 to <6 years old, and 0 to +1 D in children ≥6 years old. The IOL power was calculated by the SRK/T formula. 

Data Collection

The following data were extracted from electronic medical records: sex, ocular or systemic comorbidities, age at surgery, preoperative ocular parameters including AL, flat keratometry (K1), steep keratometry (K2), AK (calculated as [K1 + K2]/2), implanted IOL type and power, and cycloplegic refractions at 1 month, 3 months, 1 year, and subsequently every year after surgery. Refractive examinations were completed by retinoscopy in younger children, and subjective refractions were obtained in older ones. Also recorded were postoperative complications, including glaucoma, visual axis opacification (VAO), and neodymium-doped yttrium aluminum garnet (Nd:YAG) laser treatment for VAO. The preoperative ocular parameters were measured using the IOLMaster 500 (Carl Zeiss Meditec, Jena, Germany) or KM-500 keratometer (NIDEK, Aichi, Japan) and A-scan ultrasound (Quantel Medical, Cournon-d'auvergne, France). 

Refractions were converted to spherical equivalent (SE), calculated as spherical power + (cylindrical power/2). Baseline SE referred to SE measured within 3 months postoperatively. Myopic shift was calculated as the difference between SE (visit) and SE (baseline). The rate of myopic shift represented the average yearly rate of changes in refraction. 

Statistical Analysis

All statistical analyses were performed using Stata 16.0 (StataCorp, College Station, TX). Data were expressed as median (interquartile range [IQR]) for continuous variables and frequency (percentage) for categorical variables. Baseline characteristics among age groups were compared using the χ² test or one-way analysis of variance (ANOVA) test for demographic data. The trends in myopic shift and in rate of myopic shift among different subgroups within each age group were assessed by mixed-effects linear regression, accounting for the correlations between two eyes of a subject. Correlations between ocular parameters and rate of myopic shift were examined using Spearman correlation analysis. Furthermore, mixed-effects linear regression was also utilized to evaluate the potential factors influencing myopic shift and rate of myopic shift, adjusting the correlation between two eyes of a participant. Factors with P < 0.10 in univariable regression were included in the multivariable model. Two-sided P values were employed, and statistical significance was defined as P < 0.05. 

Results

Between 2010 and 2019, A total of 634 patients underwent cataract surgery and were evaluated. Out of these, 111 children (222 eyes) were eligible for the study (Fig. 1). The median age at surgery was 4.36 years (IQR, 3.16–6.0), and the median follow-up was 4.18 years (IQR, 3.48–4.64). In total, 212 eyes underwent in-the-bag IOL implantation, and 10 eyes underwent ciliary sulcus IOL implantation. Table 1 presents the participants’ clinical characteristics, myopic shift at 4 years postoperatively, and rate of myopic shift. There were no significant differences in sex, preoperative AK, or follow-up duration among age groups (all P > 0.05). Preoperative AL and baseline SE significantly differed among age groups (P = 0.01 and P < 0.001, respectively). At 4 years postoperatively, the median myopic shift was −2.25 D for the group of children 2 to <4 years old, −1.50 D for those 4 to <6 years old, and −1.31 D for those ≥6 years old. The median rate of myopic shift was −0.58 D/yr for the 2 to <4 years old group, 0.45 D/yr for the 4 to <6 years old group, and −0.40 D/yr for the ≥6 years old group. At the last visit, the median myopic shift was −2.44 D for the 2 to <4 years old group, −1.75 D for the 4 to <6 years old group, and −1.38 D for the ≥6 years old group. The median rate of myopic shift was −0.55 D/yr for the 2 to <4 years old group, 0.45 D/yr for the 4 to <6 years old group, and −0.35 D/yr for the ≥6 years old group. Figure 2 illustrates the changes in SE refraction for individuals from baseline to the last visit. Supplementary Figure S1 displays the relationship between preoperative AL and AK by age group. During the follow-up duration, no one developed ocular hypertension or glaucoma. VAO happened in 50 eyes, and 48 eyes received Nd:YAG laser treatment. 

Figure 1.

 

Flowchart of patient evaluation.

Table 1.

 

Clinical Characteristics and Myopic Shift of Participants by Age Group

 

Figure 2.

 

Loess curve showing the changes in spherical equivalent refraction for individuals from baseline to the last visit.

Trends in Myopic Shift and in Rate of Myopic Shift Among Different Subgroups Within Each Age Group

As preoperative AL increased, a decreased trend was observed in both myopic shift and rate of myopic shift across all age groups. This trend reached statistical significance in the 4 to <6 years old group (P = 0.008 and P = 0.003, respectively) and the ≥6 years old group (P = 0.002 and P < 0.001, respectively) (Table 2). However, no linear trend was observed in the relationship between preoperative AK and myopic shift or rate of myopic shift across all age groups (all P > 0.05) (Supplementary Table S1). 

Table 2.

 

Trends in Myopic Shift and in Rate of Myopic Shift Among AL Subgroups Within Each Age Group

 

Significantly negative correlations between preoperative AL and rate of myopic shift were noted in all age groups (P = 0.001, P < 0.001, and P < 0.001, respectively) (Figure 3). Conversely, no statistically significant correlation was found between preoperative AK and rate of myopic shift across all age groups (all P > 0.05) (Supplementary Fig. S2). 

Figure 3.

 

Correlation between axial length and rate of myopic shift in each age group.

Potential Factors Associated With Myopic Shift and Rate of Myopic Shift

The univariable analysis demonstrated an association between preoperative AL and myopic shift, and this relationship was still statistically significant in the multivariable analysis. Specifically, longer AL (AL above the 75th percentile) was associated with less myopic shift (β = 1.01; 95% confidence interval [CI], 0.37 to 1.64; P = 0.002) (Table 3). Shorter AL (AL below the 25th percentile) showed an association with greater myopic shift in the multivariable analysis, albeit not reaching statistical significance (β = −0.59; 95% CI, −1.20 to 0.02; P = 0.06) (Table 3). Also, younger age at surgery was significantly associated with greater myopic shift (β = 0.19; 95% CI, 0.05 to 0.33; P = 0.008) (Table 3). 

Table 3

 

Linear Regression Assessing Factors Associated With Myopic Shift

 

Similarly, preoperative AL also showed an association with rate of myopic shift. Shorter AL was associated with greater rate of myopic shift (β = −0.17; 95% CI, −0.32 to −0.02; P = 0.03), whereas longer AL was associated with less rate of myopic shift (β = 0.26; 95% CI, 0.10 to 0.41; P = 0.001) (Table 4). Additionally, younger age at surgery was significantly associated with greater rate of myopic shift (β = 0.05; 95% CI, 0.01 to 0.08; P = 0.008) (Table 4). Sex, preoperative AK, IOL implanted position, VAO, Nd:YAG laser treatment for VAO, and length of follow-up showed no association with myopic shift or rate of myopic shift (all P > 0.05) (Tables 3 and 4). 

Table 4.

 

Linear Regression Assessing Factors Associated With Rate of Myopic Shift

 

Adjustment of Target Refraction According to Preoperative AL

Based on our findings, we formulated a guideline for the adjustment of target refraction based on preoperative AL (Table 5). Our selection of cut-off values was guided by the 25th percentile and 75th percentile of AL distribution within each age group. Additionally, we also considered the convenience of these values for clinical use by surgeons as cut-off points. Shorter AL was associated with greater myopic shift, and this phenomenon was more obvious in older children (Table 2). Therefore, a greater amount of undercorrection was required for these patients. For long eyes, the amount of adjustment was decided based on the myopic shift of each longer AL subgroup and the child's age at surgery. 

Table 5.

 

Adjustment of Target Refraction According to Patients’ Preoperative AL

 

Discussion

In this longitudinal cohort study, which included 222 eyes from 111 children, data indicated that greater myopic shift occurred in pediatric patients with cataracts who underwent primary IOL implantation at a younger age. Moreover, when comparing myopic shift among AL subgroups and AK subgroups within each age group, a decreasing trend in both myopic shift and rate of myopic shift was observed with an increase in preoperative AL across all age groups. Multivariable analysis revealed that greater myopic shift and rate of myopic shift were associated with a younger age at the time of surgery. A greater rate of myopic shift was also associated with shorter AL, whereas less myopic shift and the rate of myopic shift were associated with a longer AL. However, no association was found between preoperative AK and myopic shift. 

In alignment with a multitude of preceding research,^{6,9,10,17–19} the present study observed a greater myopic shift associated with a younger age at the time of surgery in pediatric cataract patients who received primary IOL implantation. This phenomenon is widely acknowledged and factored into decision-making process when determining target refraction for pediatric cataracts in clinical practice. Based on the age at which surgery is performed, several guidelines for selecting target refraction in pediatric cataract patients have been proposed.^16,20–23 These guidelines all suggest a decrease in targeted residual hyperopia as a child's age at surgery increases. However, it is important to highlight that, due to the variability in the reported myopic shift across various studies, the recommended target refraction for children of the same age differs across these guidelines. Factors contributing to this variability include the varied sample size and follow-up duration in these studies. Long-term data and well-designed clinical studies in standardized patient populations would help address the long-term refractive prognosis of pediatric cataracts and provide a reliable reference for the selection of postoperative target refraction for pediatric cataracts. 

In the current study, children in the 2 to <4 years old group had longer ALs, and the interquartile ranges of ALs in the 4 to <6 years old group and the ≥6 years old group were wider compared with the preoperative ALs of pediatric cataract patients reported by VanderVeen et al.¹⁶ and ALs of age-matched normal children reported by Nihalani et al.²⁴ A multivariable analysis revealed an association between shorter AL and an increased myopic shift, whereas a longer AL was associated with a reduced myopic shift. This finding aligns with the results reported by Zhou et al.,¹⁴ where they observed that children undergoing cataract surgery with longer ALs showed less myopic shift compared to those with shorter ALs. Liu et al.¹³ demonstrated that shorter AL was associated with greater myopic shift in children with Marfan syndrome who underwent cataract surgery. Valera Cornejo and Flores Boza²⁵ reported the myopic shifts in a cohort of 46 patients who received cataract surgery under 4 years of age. The authors observed a trend (not reaching statistical significance, P = 0.098) to a greater myopic shift with shorter AL in the bilateral cataract group, and this trend was consistent with the results we observed in the same age group (2 to <4 years old). Considering the close relationship between axial elongation and myopia development, and in light of the evidence from our study indicating an association between preoperative AL and myopic shift, we postulate a potential association between preoperative AL and axial elongation in pediatric pseudophakic eyes. Various factors influencing axial growth in children after cataract surgery have been identified, including age at surgery, additional ocular surgery, postoperative glaucoma, and poor visual acuity.^26–30 Liu et al.¹³ reported that preoperative AL showed a negative correlation with axial elongation. Similarly, Trivedi and Wilson¹⁵ observed that eyes with shorter AL than the fellow eyes exhibited a rate of axial growth exceeding that of the eyes with longer AL. Conversely, Fan et al.¹⁷ reported that the final AL was not associated with preoperative AL. Due to the lack of patients’ AL data at each visit, the current study was unable to explore the relationship between preoperative AL and AL growth. Future research is necessary to address this question. 

Preoperative AK showed no association with myopic shift in the present study. In contrast, VanderVeen et al.¹⁶ provided evidence that a flatter cornea was associated with an increased myopic shift. This discrepancy in the findings could be attributed to the fact that methodology employed in our study for delineating subgroups is different from theirs. Moreover, it is worth noting that VanderVeen et al.¹⁶ observed a trend toward increased myopic shift in eyes with shorter AL, albeit not reaching statistical significance. We suppose that with a larger sample size this trend might achieve statistical significance. 

According to our findings, an appropriate adjustment of target refraction based on baseline AL is suggested to improve the long-term refractive outcome of children. Dahan and Druseda³¹ suggested that, in children under 2 years old, IOL power could be selected according to AL only. Trivedi et al.³² developed a model to predict children's AL into adulthood, and then they used the predicted AL in adulthood to calculate IOL power for pediatric patients. However, it would be more convenient for surgeons to employ specific AL cut-off values for such adjustments in clinical work. It is noteworthy that not all age groups exhibited consistently longer AL compared to younger age groups in our study. The reason may be the influence of cataract on children's eye development. 

Limitations of this study include the constrained number of eyes within certain age groups and the utilization of more than one device for preoperative parameter measurements. Despite these limitations, our study included a large number of pseudophakic eyes. Also, with the midterm follow-up, we found an association between preoperative AL and myopic shift in children who underwent cataract surgery, and we have proposed a guideline to adjust the target refraction based on AL. 

In conclusion, age at surgery was associated with myopic shift following primary IOL implantation in pediatric cataracts, with greater myopic shift occurring in children who underwent IOL implantation at a younger age. In eyes of all age groups, a decreasing trend in myopic shift with an increase in preoperative AL was observed. For pediatric cataracts of the same age, adjusting the target refraction based on preoperative AL could potentially improve the long-term refractive outcomes. 

Acknowledgments

Supported by grants from the Municipal Government and School (Hospital) Joint Funding Programme of Guangzhou (2023A03J0174, 2023A03J0188) and State Key Laboratories' Youth Program of China (83000-32030003). 

Disclosure: Y. Li, None; Y. Tan, None; C. Xu, None; G. Jin, None; H. Chen, None; L. Jin, None; L. Luo, None; W. Chen, None; H. Lin, None; Y. Liu, None; Z. Liu, None