Article Text
PDF
Abstract
Aim To evaluate Frisby Davis distance (FD2) stereotest for determining the timing of surgical intervention in intermittent exotropia (X(T)).
Methods A prospective case–control study was conducted including 30 patients with X(T) and 30 age-matched controls. Stereoacuity was measured preoperatively and 3 months postoperatively using FD2 for distance and TNO and Randot for near.
Results Preoperative distance stereoacuity was 43.83±35.51 arcsec (median 30 arcsec; range 10–120) which improved postoperatively to 27±33.74 arcsec (median 15 arcsec; range 5–120) (p=0.001). Cases with FD2 stereoacuity worse than 70 arcsec did not show significant improvement. Mean preoperative near stereoacuity by TNO was 94.00±79.48 arcsec (median 60 arcsec) and Randot was 50.33±39.23 arcsec (median 30 arcsec) which improved to 80.00±80.08 arcsec (median 60 arcsec) and 34.17±57.00 arcsec (median 20 arcsec), respectively, after surgery (both p=0.001). The controls had a mean distance stereoacuity of 14.66±4.13 arcsec (median 15 arcsec; range 5–20) and near stereoacuity of 63.00±21.35 arcsec (median 60 arcsec (TNO)) and 23.66±5.07 arcsec (median 20 arcsec (Randot)). There was a significant correlation between FD2 and Randot in the cases but not in controls (p=0.005), however no correlation was found between TNO and FD2.
Conclusions Distance stereoacuity is reduced in X(T) to a greater extent than the near stereoacuity and both improve after surgery. FD2 is useful for deciding timing of surgery and a stereoacuity worse than 20 arcsec is an indication for surgical intervention. A preoperative distance stereoacuity which is worse than 70 arcsec implies a poor prognosis for stereoacuity improvement after surgery.
- Muscles
Statistics from Altmetric.com
Introduction
Intermittent exotropia (X(T)) is a common form of strabismus.1 Most of the cases are progressive and often require surgical correction in view of deteriorating stereopsis.2 However, there are no well accepted guidelines regarding the timing of surgical management. The severity and stability of this disorder or the success of a therapeutic intervention assessment are entirely subjective in nature and include increasing frequency of complaints like diplopia, photophobia or monocular eye closure and clinically by poor control on cover testing. However, these have inter-observer and intra-observer variation. To make the subjective criteria more quantitative the Newcastle control score was given and it was recommended that that a high (poor) NCS (≥4) at the latest follow up indicated need for surgery.3
Since the distance control of fixation in X(T) is often the first to deteriorate, distance stereoacuity may be an early objective measure of the degree of control in X(T). It has been suggested that real depth-based FD2 gives an excellent distance stereoacuity threshold close to the true value range.4
In this study we primarily evaluated the FD2 stereotest as a tool to help in early detection of functional deterioration in X(T) and its possible recovery after surgical intervention. As a secondary objective, we also studied the effect of X(T) on near stereoacuity and examined any correlation between near and distance stereoacuity.
Subjects and methods
A prospective case–control study was conducted at a tertiary eye care hospital after permission from the institutional review board. Thirty consecutive patients with X(T), aged more than 6 years and planned for surgery were enrolled from the squint clinic. Patients with basic or divergence excess type of X(T) were included while those with convergence insufficiency type of exotropia were excluded. Other exclusion criteria included presence of amblyopia, more than 2 prism dioptres of vertical deviation, history of previous squint surgery, high refractive error (more than 4 D spherical or 2 D cylinder) and any other ocular pathology likely to reduce vision and stereoacuity. Subjects in the control group were age matched and included those without any manifest squint or any confounding ocular illness likely to affect vision and stereoacuity.
Each of the cases and controls underwent a detailed history and ocular examination, including a cycloplegic refraction, to ascertain their inclusion in the study. All tests were performed after appropriate refractive correction. For the cases, a prism bar cover test with an accommodative target was performed to measure the angle of deviation for both near (33 cm) and distance (6 m) in primary position, up gaze and down gaze. As a routine workup, patch testing was done for all X(T) cases and AC/A ratio was calculated.
All subjects underwent distance stereoacuity testing using the Frisby Davis distance (FD2) stereotest. First subjects were sensitised to the test by making them stand at 1 m distance and appreciate the maximum disparity between the test objects. This was followed by formal testing at 6 m while care was taken to obviate any monocular cues.5 ,6 All patients responded satisfactorily at 6 m distance and these values defined distance stereoacuity. In the case of manifest exotropia for distance (such as in patients who were constantly exotropic for distance but orthotropic or exophoric for near), the FD2 stereotest was performed after correction of distance deviation by prisms and ensuring fusion for the 6 m targets.
Near stereoacuity was measured by the TNO stereotest using red green anaglyphic filter glasses at 40 cm while holding the booklet perpendicular to the subject's visual axis. After the screening plates were successfully completed, the graded plates were presented until the subject could no longer appreciate the three-dimensional shape correctly. For the Randot stereotest, the booklet was viewed through polarising glasses while being held at 40 cm. The stereopsis on Randot test was quantified using the circles on a random dot background. On each test, the lowest disparity that the subject was able to detect was recorded as the stereoacuity in seconds of arc.
Each of these tests were performed preoperatively and postoperatively at 1-week, 1-month and 3-month follow-up. In the control group, distance and near stereoacuity were measured in a similar fashion. For the purposes of analysis, the stereoacuity at postoperative 3-month follow-up was used.
The surgery performed in the cases included either a unilateral or bilateral lateral rectus recession or a unilateral lateral rectus recession and medial rectus resection.
The analysis was done using appropriate non-parametric tests on SPSS V.15.0 statistical software (SPSS Inc, Chicago, Illinois, USA). Descriptive statistics included mean, median, SD and range for each parameter. Intergroup comparisons were done using appropriate non-parametric tests (Mann–Whitney U and Wilcoxon rank sum tests). Spearman's correlation was used for evaluating associations between parameters. For comparative analysis, the parameters of the final follow-up visit at 3 months postoperatively were used.
Results
The demographic and clinical profiles of the study and control groups are shown in table 1.
Table depicting demographic and clinical characteristics
While the study group had a female preponderance, the age distribution was similar in both groups. The refractive errors in both groups were similar, with the majority of the subjects having emmetropia. Low refractive errors were present in four cases in the study group (+0.5 to − D (spherical equivalent)) and five cases in the control group (+0.5 to −2.5 D (spherical equivalent)). The subjects with exotropia had significantly worse near and distance stereoacuity compared with the control group (table 2).
Table depicting ocular deviations and stereoacuity parameters for cases and controls with corresponding p values
The surgical intervention included unilateral lateral rectus recession in 8 (26.66%) patients, unilateral lateral rectus recession and medial rectus resection in 13 (43.33%) patients and bilateral lateral rectus recession in 9 (30%) patients. The mean preoperative distance deviation in subjects who underwent single muscle surgery was 22.60±3.20 prism dioptres compared with 30.68±6.30 in those who underwent two-muscle surgery.
The ocular deviation and stereoacuity parameters are depicted in table 2. There was a significant improvement in the distance and near deviations and 28 cases had a successful alignment after surgery (orthotropia or residual exotropia <8Δ or consecutive esotropia <5Δ). Two cases which were not aligned at 3 months postoperatively were planned for resurgery. The improvement in distance and near stereoacuity was observed immediately at 1-week follow-up and was not significantly different from that at 3-month postoperative follow-up.
The study group had a median distance stereoacuity (FD2) of 30 arcsec (range 10–120) preoperatively which improved postoperatively to a median of 15 arcsec (range 5–120) at 3 months (p>0.05). Nearly the entire change in distance stereoacuity occurred at 1 week after surgery. Median distance stereoacuity in the control group was 15 arcsec (range 5–20).
All cases having preoperative distance stereoacuity up to 70 arcsec showed improvement after surgery of which five cases had a normal preoperative distance stereoacuity (<20 arcsec).
To avoid the potential bias of stereoacuity (being unable to improve if the preoperative stereoacuity was already normal), an additional analysis was performed that included only patients with subnormal stereoacuity. Twenty-five of 30 patients (83.33%) had sub-normal distance stereoacuity (worse than 20 arcsec) preoperatively and in this subgroup, distance stereoacuity improved from a median of 30 arcsec (mean 50.20±35.63) to 15 arcsec (mean 30.40±36.11) (p=0.005). Twenty of these 25 patients (80%) achieved normal stereoacuity postoperatively. However, five patients (each with stereoacuity worse than 70 arcsec in the preoperative period) did not show a significant improvement after surgery and stereoacuity continued to remain subnormal (p=0.09).
A subgroup analysis comparing the group of patients having preoperative distance stereoacuity better and worse than 70 arcsec showed a significant difference between the groups in the amount of change of stereoacuity and final postoperative stereoacuity (p=0.007 and p=0.001, respectively). Mean distance stereoacuity at 3-month follow-up in cases did not reach that of the control groups.
The near stereoacuity measured using the Randot in the study group had a median value of 30 arcsec (range 20–200 arcsec) with 10 of the 30 patients having a subnormal near stereoacuity. Postoperatively (at 3 months) significant improvement (p=0.005) was seen in this stereoacuity, which came to a median value of 20 arcsec (range 20–200 arcsec). Three patients did not show any improvement in the near stereoacuity after surgery. In the control group, the median stereoacuity by Randot was 20 arcsec (range 20–40 arcsec). Despite a significant improvement in near Randot stereoacuity after surgery, it was statistically worse than the control group (p=0.03).
In the study group, the TNO near stereoacuity had a median value of 60 arcsec (range 30–480 arcsec) in the preoperative period. Postoperatively (at 3 months), the stereoacuity showed significant improvement to a median value of 60 arcsec (range 30–480) (p=0.009). One case did not show an improvement in near stereoacuity after surgery. In the control group, the median stereoacuity was 60 arcsec (range 30–120). Despite significant improvement in near TNO stereoacuity after surgery, it was still worse than the control group (p=0.03). There was no significant improvement in near stereoacuity in cases which failed to improve by FD2.
There was a significant correlation (p=0.005) between FD2 and Randot in preoperative and postoperative cases but not in the control group. No significant correlation was found between TNO and FD2 in the study group in the preoperative and postoperative period or in the controls (table 3).
Table depicting significant correlations between various stereotests in the preoperative and postoperative period
Discussion
X(T) is known to affect distance and near stereoacuity and there is consequent improvement in these after surgery and binocular alignment.7–12 Previous studies evaluating the FD2 stereotest have demonstrated significant reduction in X(T) which has shown improvement after surgery.7–11 Previous studies by the authors have shown improvement in distance (FD2) and near (TNO and Randot) stereoacuity after surgery for X(T), though they did not reach normal levels even 6 months postoperatively.7 ,8 Contrary to our study, Adams et al did not find a significant improvement in near stereoacuity (near FD2 and preschool Randot) after surgery for X(T) despite an improvement in distance stereoacuity.9 Yildirim et al proved that the distance stereoacuity is affected to a much greater extent than near stereoacuity, which may be at near normal levels in cases of X(T).10 While they stated that optimal surgical alignment of the eyes had a bearing on stereoacuity improvement, we found that it was the extent of reduced preoperative stereoacuity which mattered and this concurs with previous literature.7 ,8This study builds upon the previous findings of the authors and defines a precise cutoff of distance stereoacuity to determine the timing for surgery.
The FD2 stereotest was chosen as the method of testing stereoacuity as it is a real depth stereotest and it has been proven that the real world contour-based FD2 gives an excellent distance stereoacuity threshold close to the true values in X(T).4 The importance of an accurate distance stereoacuity also stems from the fact that the distance control of fixation in X(T) is often the first to deteriorate and the distance stereoacuity is an early objective measure of the degree of control in intermittent exotropia.10–12
In the present study, distance stereoacuity was measured by FD2 at the furthest distance of 6 m to attain the most accurate measurement and earliest change of stereoacuity in concordance with previous study protocols.5 ,7 ,9 Further, all conditions which may lead to inaccurate results such as the presence of dissociated horizontal deviations or high refractive errors were excluded for optimum results.13 However, there is a potential limitation regarding patients who had a manifest deviation for distance and required prisms to achieve alignment before attempting the FD2 stereotest. This stems from the fact that prisms have been shown to adversely affect stereoacuity14 However, the prisms have been shown to negatively impact stereotesting for TNO and Titmus tests, both of which require dissociating glasses and such an impact may possibly be minimal on the FD2 real depth test. Since the exact impact of prisms on FD2 is not clearly known, we consider it a potential limitation. Considering that a certain subset of patients with X(T) are manifest for distance and can fuse for near, prisms had to be used for measuring preoperative FD2 stereoacuity to prevent exclusion of such patients and a consequent comprehensive study result.
X(T) cases present with variable levels of distance stereoacuity as is elucidated in our series in which 5 cases had normal and 25 cases had subnormal stereoacuity, 4 of which improved after surgery. The cases which did show improvement in distance stereoacuity after surgery were those with a preoperative stereoacuity better than 70 arcsec. This has a prognostic significance and we suggest this as a cutoff beyond which only cosmetic correction, but no functional gain, is expected. The non-improvement of distance stereoacuity can be attributed to central suppression.8 ,15
We found that the normal range of distance stereoacuity level by FD2 in the control group was 5–20 arcsec, thus defining the normative values of distance stereoacuity. Our study has also generated a normative database for near stereoacuity (TNO and Randot) in the 30 controls, which further validates the findings of our previous studies.7 ,8
In contrast to the nearly normal near stereoacuity described in studies on children with X(T) of short follow-up, the near stereoacuity levels at presentation were poor in our cases.16 This could be due to the late presentation of cases as the patients normally present only after they become symptomatic, with the deviation becoming manifest for longer durations.17 It is shown that very longstanding exodeviation causes progressive decrease in stereoacuity. In fact, deterioration in stereoacuity is one of the manifestations of progression of X(T).18 Initially the distance stereoacuity deteriorates but as the disease progresses, near stereoacuity also worsens.6
In this study, we found significant correlation between FD2 and Randot. Thus it would be helpful to use FD2 for distance and Randot for near stereoacuity simultaneously in preoperative follow-up of intermittent exotropia cases. The number of cases with subnormal near Randot and FD2 were too few to enable a meaningful analysis of the additional prognostic value of near Randot to FD2 and a larger study would be required. We could not prove such a correlation in the control group, probably due to the fact that both the distance and near stereoacuity values were within the normal limits with a narrow range. A larger population of controls would be needed to prove such an association.
One limitation of this study is that we included patients aged over 6 years and so the findings may not be directly applied to children under 6 years. However, a similar outcome is also likely for younger children.
Surgery in X(T) may be indicated if the distance stereoacuity by FD2 is worse than normal control groups. A cutoff may be kept at 20 arcsec, which is an indicator of loss of fusional control or functional deterioration. Functional improvement is possible in cases having distance stereoacuity up to 70 arcsec. However, cases having very poor (>70 arcsec) stereoacuity are less likely to have improvement in stereopsis after surgery and may get only cosmetic correction. We recommend surgery optimally timed with the FD2 stereoacuity assessment. Near stereoacuity is also affected in X(T) and shows some recovery after surgery. Randot has a moderate strength of correlation with FD2 and may also be a potential tool for deciding timing of surgical intervention.
To conclude, FD2 is a useful test for measuring distance stereoacuity and helpful in determining timing of surgery to prevent loss of stereopsis and also to restore the loss, if treated early.
References
Footnotes
-
Contributors Design and conduct of study: PS, VM, AS; collection of data: AS, PS, RS; analysis and interpretation of data: AS, DS, PS, AS; preparation of manuscript: PS, AS, DS, AS; review and approval of manuscript: VM, PS, RS, DS.
-
Competing interests None.
-
Ethics approval The study was conducted in compliance with the Declaration of Helsinki, and Institutional Ethics subcommittee, All India Institute of Medical Sciences, New Delhi, India.
-
Provenance and peer review Not commissioned; externally peer reviewed.