DOI: https://doi.org/10.21203/rs.3.rs-1941993/v1
Intermittent exotropia (IXT) is a relatively common childhood strabismus worldwide, affecting 0.5~1% of the general population 1,2. Patients with this condition typically present with an outward deviation of one or both eyes when fatigue, poor concentration, or fixating the far objects.
A whole range of treatment options is available, including watchful waiting, occlusion therapy, overminus lens therapy, orthoptic exercises, and surgery. However, surgery still carries risks, such as recurrence and overcorrection, especially overcorrection which may adversely affect stereopsis, and in addition, there is no consensus on the optimal timing of surgery.
Non-surgical treatments are sometimes administered to prevent IXT from deteriorating, which has not been widely accepted by physicians because of the uncertainty of its therapeutic effects. For example, Brian G. Mohney found that deterioration over 6 months was uncommon among children 12-35 months of age with previously untreated IXT with or without occlusion therapy3. Mohammad Reza Akbari noted that compared with observations, part-time occlusion therapy appeared to be effective in improving deviation control in 3 to 8-year-old children with intermittent exotropia based on two common office-based control scales4.
Overminus lens therapy and occlusion therapy for intermittent exotropia are frequently prescribed non-surgical treatment modalities. Recently, several randomized controlled trials (RCTs) have directly compared these two treatment modalities with observation (watchful waiting), with mixed results3-9. A meta-analysis is necessary to better understand the effectiveness of overminus lens and occlusion therapy in the treatment of IXT.
This follow-up meta-analysis was aimed to assess the overall efficacy of the commonly used non-surgical treatment modalities (overminus lens therapy and part-time occlusion therapy) in slowing the progression of IXT in children, and the change in deviation control was chosen as the first outcome measure. It was also investigated whether there was a difference in treatment efficacy between overminus lens and occlusion therapy. To exclude the effect of confounding factors, only high-quality RCTs were included in this meta-analysis.
Search strategy. This meta-analysis was based on Cochrane Review Methods. Two researchers (DSS and JQ) independently searched the following electronic databases and screened relevant literature up to July 2022: Web of Science, Medline, Embase and the Cochrane Register of Controlled Trials data bases. The searched keywords were as follows: “Patch* occlu*,” “minus lens”, “overminus*”, “overcorrecting minus lens”, “minus overcorrection”, “overminus”, “therapy,” “treatment outcome,” “exotropia,” “exodeviation,” “divergent strabismus,” “external strabismus,”, and “divergent squint”. The search language was not limited to English, and for non-English articles, English abstracts were used. Furthermore, references to all searched articles were also manually searched for any additional relevant publications.
Study inclusion and exclusion criteria. The controlled studies met the following inclusion criteria: (1) indicator of treatment outcome included exotropia control change at distance or at near; (2) subjects in the trials suffered from intermittent exotropia without prior history of surgical or nonsurgical treatment other than refractive correction; (3) Nonsurgical treatment (including occlusion therapy and overminus lens therapy) and observation were compared in a comparative study; (4) there was at least a 8-week follow-up period in the study; and (5) the type of study included was randomized controlled trial (RCT). In addition, study was excluded based on two exclusion criteria: (1) Letters, correspondence and reviews; (2) unpublished articles (e.g., conference abstracts), case reports or case series without control group.
Study selection and data extraction. The literatures were independently retrieved and screened by two investigators, DS Song, J Qian. The two researchers extracted the data provided in the articles using a literature data extraction table.
If any disagreement was encountered, the third author (ZJ Chen) decided on the disagreement. Information from included literatures was extracted as followed: first author, publication year, length of follow-up, type of intervention, distance control at baseline, sample size, and OML power/Occlusion strategy. Statistical analysis was carried out by using data from the last visit. This meta-analysis followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.
Quality assessment. The quality of the Randomized controlled trials (RCTs) was evaluated using the Cochrane evidence-based medicine system10. The quality assessment for each trial included five domains: Election bias (randomization order generation, allocation hiding); implementation bias (blind method); measurement bias (blind method in outcome evaluation); loss to follow-up bias (incompletely resolved data); publication bias (selective reporting of research results); and other biases.
Statistical analysis. Statistical analysis was conducted using Review Manager 5.3 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2008). Weighted mean differences (WMD) and 95% confidence interval were calculated. Heterogeneity between studies was assessed using I2 and Q tests11, and the fixed effect model was performed when P＞0.05 or I2＜50%, otherwise the random effect model was employed. Pooled results with a p-values < 0.05 were considered statistically significant.
Figure 1 showed the flow chart illustrating our study procedures. Through an electronic literature search, it was found that one thousand and twenty-two studies were identified. After screening, seven RCT studies met the inclusion criteria and identified 1,046 patients of which 527 were included in the non-surgical treatment group and 519 were included in the observation group. Three studies compared the effect of minus lens therapy to observation, three studies compared the effects of part-time occlusion therapy with observation, and one study compared the effect of overminus lens, part-time occlusion, and observation. A summary of these included studies is shown in Table 1. Two authors independently assessed the risk of bias in each included study (Figure 2).
Change in distance exotropia control. All included studies reported changes in distance control of exotropia. In the commonly used non-surgical treatment compared with the observation group, four studies reported data on overminus lens therapy. There were also four studies reporting occlusion therapy. The pooled data showed that the decrease was significantly greater in the non-surgical treatment group than in the observation group (MD=-0.51, 95% CI: -0.66 to -0.36, P<0.001). In addition, OML showed significantly better efficacy than PTO in reducing distance deviation control score of intermittent exotropia, with statistically significant differences between subgroups (P=0.03).
It was found that there was significant heterogeneity in the treatment effects of the overminus lens, and sensitivity analysis (removal of study 1 by 1) was performed to examine the sources of heterogeneity. By omitting the study by Jit et al, heterogeneity decreased from 93% to 0%, with no reversal of the overall effect (Figure 3).
Change in near exotropia control. Six studies reported changes in exotropia control at near between the non-surgical treatment group and the observation group. At near, the WMD for change in exotropia control between the two groups was -0.54 (95%, -0.67 to -0.41; P < 0.001). PTO was more effective compared with OML in reducing near IXT control score(P=0.003).
There was statistical heterogeneity between the groups (P < 0.001, I2 = 89%), which was reduced from 81% to 46% by omitting the study of Jit et al. and the overall effect was not reversed (Figure 4).
Change in deviation magnitude at distance and at near. Five studies addressed changes in the magnitude of exodeviation angles at distance and near. Two studies compared the effect of PTO with observation, and three studies compared the effect of OML with observation. Pooled results showed that for intermittent exotropia, non-surgical treatment was more effective in reducing distance and near angles compared with observation (MD=-2.88, 95% CI: -3.75 to -2.00, P<0.001 at distance; MD=-3.24, 95% CI: -3.27 to -2.21, P<0.001 at near).
A subgroup analysis was performed according to the type of non-surgical treatment. Pooled results showed that OML exhibited a better significant treatment effect than PTO in reducing both distance and near angles of deviation (P=0.02; P< 0.001).
There was no significant between-study heterogeneity in the change in the magnitude of deviation angle at near. By omitting the study by Jit et al, the heterogeneity at distance was reduced from 77% to 0%, and the overall effect was not reversed, as shown in Figure 5.
Stereopsis at near. Five studies reported near stereoacuity after non-surgical treatment or observation. The combined results showed a WMD of 0.01,95% CI (-0.04, 0.05), which indicates that there was no significant difference in near stereoacuity between the two groups, as shown in Figure 6. No statistical heterogeneity was found between the two groups (P = 0.71, I2 = 0%).
Myopic shift. The myopic shift was reported in two out of 4 articles focused on a comparison between OML and Observation. The combined results showed that the WMD in the myopic shift between the overminus group and observation group was -0.32D (95%CI, -0.43 to -0.21D; P<0.001), there was a greater myopic shift in the OML group in comparison to the observation group, as shown in Figure 7.
An analysis was conducted to examine the impact of commonly used non-surgical treatment modalities on intermittent exotropia. Seven published RCTs were included in this analysis. Of these, four studies reported data on overminus lens therapy, and four studies reported occlusion therapy. The pooled results showed that the non-surgical treatment demonstrated higher outcomes compared to observation, with greater reductions in exotropia control and angle of deviation at both distance and near. Similar results were found when comparing post-treatment near stereoacuity. Further analysis revealed that OML therapy showed a better apparent treatment effect than PTO therapy.
Surgery is the most common treatment modality for intermittent exotropia as it is directly effective in reducing the exotropia angle and promoting motor control, however, surgery still carries risks. The advantages of non-surgical treatments include no significant side effects and ease of perform. At the same time, there are some concerns with non-surgical treatment therapies, such as a significant psychological burden on the family12, skin irritation in patients with sensitive skin 5, and unsatisfactory treatment results. In addition, it does not result in a significant reduction in exotropia deviation. Therefore, the clinical application of non-surgical treatment is indeed limited in this way. By control, it is meant the frequency of manifest deviations and the ease of readjustment. In the last few years, there have been many reports exploring the impact of non-surgical treatments on control, with mixed results. In our meta-analysis, patients with intermittent exotropia benefited more from non-surgical treatment.
It should be noted, however, that there was significant heterogeneity between the two groups when comparing changes in exotropia control and distance deviation angle. In the random-effects model, heterogeneity remained evident. A sensitivity analysis was conducted to look for sources of heterogeneity and we found the study by Jit et al. to be the main source of heterogeneity7. The study focused on comparing overminus lenses therapy with observation for intermittent exotropia and developed a practical algorithm. Its ability to individually tailor the overminus lens requirements for treating exotropia was demonstrated by comparison with a fixed 2.50D overminus lens strategy, with over 74% of participants requiring a stronger overminus lens, and the efficacy of OML treatment in this study was greater than that found in the other included studies. These findings may be partly attributable to the greater efficacy of bespoke OML in treating IXT compared to standard fixed overminus power.
Part-time occlusion has been used as an anti-suppression treatment for the non-surgical management of intermittent exotropia with variable efficacy. Akbari et al found that patching therapy resulted in significant improvements in distance deviation control by using a 3-point system and in near deviation control by using a 6-point system, no significant improvement was observed in near control according to the 3-point scale. They also observed a significant improvement in near stereopsis13. A randomized controlled trial including 358 untreated patients aged 3-11 years with intermittent exotropia showed that after 6 months, the memory of the patching was slightly better than observed in terms of control scores in the near and this difference was also found in terms of control in the distance, however, it did not reach statistical significance. In addition, they did not observe a significant improvement in near stereopsis5. Alkahmous et al. evaluated the effect of occlusion treatment in 21 patients with intermittent exotropia aged 4-10 years. Using a 6-point scale, they found that 77.7% of occlusion treatment cases were successful in distance deviation control and 100% were successful in near deviation control. However, the occlusion treatment did not significantly reduce exodeviation angle14. On the other hand, Brian G. Mohney found that deterioration beyond 6 months was uncommon in children between 12 and 35 months of age, with or without occlusion treatment 3.
Overminus lenses have been suggested to improve the control ability of exodeviation by increasing accommodative convergence. Overall, most researchers agree about the positive therapeutic effects of overminus lens. A retrospective study by Bayramlar et al. assessing the effectiveness of overminus lens in patients with intermittent exotropia showed that according to the Newcastle Control System (NCS) and Jampolsky's assessment, 84% of these patients achieved well control scores after a mean follow-up of 18 months, and the distant angle decreased from 25 prism diopters (PD) at baseline to 18 PD after overminus lens treatment (p = 0.002)15. Another study included 24 patients aged 3-11 years with intermittent exotropia and reported that overminus lens treatment significantly reduced the distance exotropia angle and improved the control leverage 16. A retrospective study of 21 children aged 1-9 years with IXT found a 5-year improvement rate of 51% with overminus lens therapy using the NCS system 17. Kaveh Abri Aghdam reported that in a retrospective study of 163 patients with IXT, 109 patients (66.8%), based on Jampolsky's qualitative assessment method, attained good controlled IXT or orthotropia after 1 year of overminus lens treatment18.
Our meta-analysis found that both non-surgical treatments were effective in reducing near and distance control scores and in reducing the angle of exodeviaton at distance, and that overminus lenses also reduced the angle of exodeviation at near, whereas part-time occlusion did not, and neither improved stereopsis at near. Overminus lenses were superior to part-time occlusion in terms of reducing the angle of deviation and increasing the deviation control. Similar to a previous study demonstrating that overminus spectacle therapy was more effective than PTO therapy in controlling IXT, the authors emphasize that poor compliance with PTO therapy reduces its efficacy8.
Notably, myopic shift due to overcorrecting minus lenses is one of the most frequently expressed concerns of clinicians in patients with IXT. A retrospective study of 153 intermittent exotropia patients comprising 74 patients treated with overminus lenses for at least 6 months,34 patients treated with overminus lenses for 5 years and 45 patients who did not receive overminus lenses therapy found the mean myopic shift was -1.52±1.80D in the 6-month treatment group, -1.54±1.80D in the 5-year treatment group, and -1.40±2.80 D in the control group after 5 years follow up, the difference in myopic shift did not reach statistical significance19. F J Rowe identified 21 patients aged 1-9 years diagnosed with intermittent exotropia in whom they received overminus lenses therapy. Consistent with the results of previous studies, myopia shift over a median of 3 years was -0.61 ± 0.87D, reflecting the normal developmental changes of refractive error through time17. However, the sample size of the above studies was limited, and the quality was not high. More recently, two high-quality, well-designed randomized controlled studies found that overcorrection of minus lenses induced significantly higher levels of myopia than observed6,7. The results of our meta-analysis showed a WMD of -0.32 D (95% CI, -0.43 to -0.21 D) for a myopic shift in the overminus groups. Jit B. Ale Magar found that myopic individuals were associated with a greater risk of myopic shift. Special care should be taken when prescribing overminus lens treatment for IXT patients with myopic.
One randomized clinical trial was excluded from our meta-analysis because the patients in that study were treated with overcorrecting minus lenses in combination with prismatic lenses. The results of this combined treatment were better than all included studies, with mean external strabismus control scores of 1.75 ± 1.18 and 5.72 ± 1.28 (95% CI: -4.63 to -3.33) in the treatment and observation groups, and mean near stereopsis of 1.91 ± 0.26 log arcsec and 2.16 ± 0.42 log arcsec (95% CI: - 0.44 to -0.06) at last follow-up. Although this study was not included in our meta-analysis, we believe it would be valuable in helping us explore the efficacy of OML combined with other non-surgical treatment modalities in the treatment of IXT20.
In our meta-analysis, change in control was selected as the primary measure of treatment outcome. Based on previous studies, non-surgical treatment appears to be less effective in reducing the exodeviation angle, control ability is a more appropriate marker for predicting efficacy. On the other hand, if control scores, angle of exodeviation, and stereopsis at the last visit are chosen as the main comparators, results may be influenced by baseline values. For example, in Merna's study, baseline control scores were much lower in the observation group than in the non-surgical group, which directly led to lower post-treatment control score in the observation group than in the non-surgical group, however, the latter was superior to the former in terms of the amount of change in control score8. Ultimately, the change in deviation control at distance and near and the change in the angle of exodeviation at distance and near were chosen as the main indicators of comparison in our meta-analysis.
In our current study, there are several flaws and shortcomings. First, even though the sample size here is the largest and the quality of the included literature is the highest, the limited number of subjects included in this meta-analysis may have contributed to some degree of selection bias and information bias. In addition, the different overminus lens power or occlusion strategies in the included studies may have contributed to bias. Secondly, given the limited data from the included studies, our meta-analysis focused on clinical outcomes at the last follow-up rather than data collected at early, mid-term and long-term follow-ups, which may have had an impact on the current results. More detailed analyses (e.g., further subgroup analyses by follow-up period) would be required to address potential biases if sufficient clinical data were available. Finally, no additional comparisons of stereotaxic changes were performed due to the lack of relevant trial data.
In conclusion, the results of this meta-analysis suggest that the commonly used non-surgical treatment modalities (overminus lens therapy and part-time occlusion treatment) resulted in a greater decrease in near and distance exodeviation control and angles in patients with intermittent exotropia. Our findings also provide clear evidence that overminus lens therapy is superior to part-time occlusion therapy reducing the angle of deviation and increasing the deviation control. However, overminus lens therapy may increase the risk of myopic shift. Clearly, many issues require further research. For example, lack of a uniform overminus lens power or occlusion strategy. In our meta-analysis, four studies focused on a comparison of OML and observed, using four different overminus lens powers, and four studies focused on a comparison of PTO and observation, using three different occlusion strategies, which may increase clinical heterogeneity. Specifically, more studies should be conducted to explore the relationship between refractive change and OML wear power or duration, whether the benefit of PTO treatment is stable after treatment termination, the relationship between compliance and treatment outcome, the relationship between myopic shift and baseline refraction, and the effect of increasing overminus lens power or occlusion duration on outcomes.
Data availability
The datasets generated during and/or analysed during the current study are available throughout the manuscript.
Author contributions
D.S. and Z.J. conceived the study. J.Q. designed and performed the searches. D.S. and J.Q. conducted data extraction, appraisal, synthesis and wrote the manuscript. All authors participated in discussions about eligibility and quality of the included studies, critically reviewed and approved the final version of the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
Corresponding author
Correspondence to Desheng Song.
Table 1. Characteristics of the studies included in the meta-analysis
|
Study |
follow-up |
Age, years |
Intervention |
Control, point |
Subjects |
OML power(D)/Occlusion strategy |
|
Magar 2021 [9] |
≥6 months |
6.8 ± 2.5 |
OML |
OML 3.8 ± 0.8 OBS 3.5 ± 0.8 |
BMR77 BTX 66 |
3.7±1.2D(-1.00~-6.25D) |
|
PEDIG 2016 [6] |
8 weeks |
OML 5.1 (1.2) OBS 5.2 (1) |
OML |
OML 3.2 (1.1) OBS 3.2 (1.1) |
OML 20 OBS 25 |
-2.50D |
|
Angela 2021 [8] |
12months |
OML 6.4(2.0) OBS 6.2(2.2) |
OML |
OML 3.2 (1.0) OBS 3.2 (1.0) |
OML 196 OBS 190 |
-2.50D(12 months), then−1.25D(3 months) |
|
Merna 2018 [7] |
≥4 months |
4.675±2.553 |
OML PTO |
OML 3.95±0.76 PTO 3.95±1.10 OBS 3.05±1.10 |
OML 20 PTO 15 OBS 25 |
-3.00D / alternately patched 4 hours daily |
|
Akbari 2020 [4] |
6 months |
PTO 5.34 ± 1.26 OBS 4.68 ± 1.33 |
PTO |
PTO 2.8 ± 1.1 OBS 2.6 ± 1.1 |
PTO 35 OBS 41 |
alternately patched 2 hours daily, In the presence of significant dominancy, the dominant eye patched 5 days a week and other eye patched 2 days a week. |
|
PEDIG 2015 [3] |
6 months |
PTO 5.24.9 (6.8) BTX 24.2 (6.5) |
PTO |
PTO 2.3 (1.1) OBS 2.5 (1.2) |
BMR87 BTX 90 |
3 hours daily for 5 months, with a 1-month washout period of no patching |
|
PEDIG 2014 [5] |
>6 months |
PTO 5.9 (2.0) OBS 6.1 (2.0) |
PTO |
PTO 2.3 (1.2) OBS 2.4 (1.2) |
PTO 159 OBS 165 |
3 hours daily for 5 months, with a 1-month washout period of no patching |
PEDIG= Pediatric Eye Disease Investigator Group; OML= overminus lens; PTO= part-time occlusion; OBS= observation; D=diopter; PD = prism diopters