Purpose
To determine whether stereopsis of infants treated for monocular cataracts varies with the type of optical correction used.
Design
Randomized prospective clinical trial.
Methods
The Infant Aphakia Treatment Study randomized 114 patients with unilateral cataracts at age 1–7 months to either primary intraocular lens (IOL) or contact lens correction. At 4.5 years of age a masked examiner assessed stereopsis on these patients using 3 different tests: (1) Frisby; (2) Randot Preschool; and (3) Titmus Fly.
Results
Twenty-eight patients (25%) had a positive response to at least 1 of the stereopsis tests. There was no statistically significant difference in stereopsis between the 2 treatment groups: Frisby (contact lens, 6 [11%]; IOL, 7 [13%]; P = .99), Randot (contact lens, 3 [6%]; IOL, 1 [2%]; P = .62), or Titmus (contact lens, 8 [15%]; IOL, 13 [23%]; P = .34). The median age at surgery for patients with stereopsis was younger than for those without stereopsis (1.2 vs 2.4 months; P = .002). The median visual acuity for patients with stereopsis was better than for those without stereopsis (20/40 vs 20/252; P = .0003).
Conclusion
The type of optical correction did not influence stereopsis outcomes. However, 2 other factors did: age at surgery and visual acuity in the treated eye at age 4.5 years. Early surgery for unilateral congenital cataract and the presence of visual acuity better than or equal to 20/40 appear to be more important than the type of initial optical correction used for the development of stereopsis.
The ability to perceive depth based on image disparity between the 2 eyes (stereopsis) is the hallmark of human binocular vision. Patients with a unilateral congenital cataract disabling enough to require surgery do not generally demonstrate stereopsis despite best efforts at visual rehabilitation. Infants who have undergone surgery for a unilateral cataract do not experience the typical level of equal and simultaneous visual input to both eyes, which is required for the development of critical cortical connections underlying stereoscopic vision. Anecdotal accounts of good stereopsis following unilateral congenital cataract surgery have been reported, but treatment regimens and stereopsis testing were not standardized. The Infant Aphakia Treatment Study randomized infants who had surgery for a unilateral cataract between 1 and 7 months of age to either primary intraocular lens or contact lens correction. This paper examines differences in stereopsis at 4.5 years of age between the 2 treatment groups.
Methods
The Infant Aphakia Treatment Study is a randomized clinical trial involving 12 sites supported by a cooperative agreement with the National Eye Institute of the National Institutes of Health. The study design, surgical techniques, patching and optical correction regimens, evaluation methods, patient characteristics at baseline, monocular visual acuity at age 4.5 years, and clinical findings at age 5 years have been reported previously. Only the elements pertinent to the current paper will be briefly described here. The study and data collection were carried out with approval from the appropriate Institutional Review Board at each site. Informed consent for the research was obtained from the parents of the participants, and the study is in accordance with the Health Insurance Portability and Accountability regulations. The off-label research use of the Acrysof SN60AT and MA60AC intraocular lenses (Alcon Laboratories, Fort Worth, Texas, USA) is covered by US Food and Drug Administration investigational device exemption G020021.
Optical Correction
Patients randomized to the contact lens group were fitted with a Silsoft (Bausch + Lomb, Rochester, New York, USA) or a rigid gas-permeable contact lens within a week after surgery. The power of the lens included a 2.0 diopter (D) overcorrection to provide a near point focus. The intraocular lens power for infants randomized to that group was calculated based on the Holladay 1 formula targeting an 8 D undercorrection for infants 4–6 weeks of age and a 6 D undercorrection for infants older than 6 weeks. For this group, a 2.0 D overcorrection was achieved by prescribing spectacles, beginning as early as the 1-month postoperative visit, when any of the following conditions existed in the treated eye: hyperopia >1.0 D, myopia >3.0 D, or astigmatism >1.5 D. At 2 years of age, all patients received bifocal glasses with near correction and were asked to wear them at all times while awake. These spectacles included distance correction for patients in the intraocular lens group. Aphakic patients continued to wear a contact lens for their distance refractive correction.
Patching Regimen
Starting at the second postoperative week, parents were instructed to have their child wear an adhesive occlusive patch over the unoperated eye for 1 hour/day for each month of age until age 8 months. Thereafter, patching was prescribed for one half of waking hours every day or all waking hours every other day, achieving half-time patching. This regimen was continued until the patients exited from the study at 5 years of age.
Visual Acuity Assessment
Monocular optotype acuity was assessed at age 4.5 years (window +1 month) by a masked traveling examiner using the Amblyopia Treatment Study (ATS)-HOTV test. Patients were tested wearing their best correction (updated at their last study visit 3 months earlier). Visual acuity was tested first in the aphakic/pseudophakic eye. The eye not being tested was occluded using a translucent occluder mounted in child sunglass frames (Good-Lite; Elgin, Illinois, USA) to minimize the amplitude of latent nystagmus under monocular conditions. The initial testing distance was 3 meters. If the child was unable to see the HOTV letters, this distance was decreased to 1 meter. If the child still could not identify the letters, the Low Vision Card (Teller Acuity Card 0.32 cy/cm) was used to test for pattern vision . If the child did not respond to the Low Vision Card, the eye was assessed for light perception using a penlight in a room with no overhead lighting. If the child did not respond to the penlight, the vision was categorized as no light perception.
Stereopsis Testing
Stereopsis was evaluated at 4.5 years of age (window +1 month) by a masked traveling examiner using 3 different tests: (1) Frisby Stereotest (Richmond Products, Albuquerque, New Mexico, USA); (2) Randot Preschool Stereoacuity test (Stereo Optical, Chicago, Illinois, USA); and (3) Titmus Fly test (Stereo Optical, Chicago, Illinois, USA).
The Frisby Stereotest consists of a series of square plates with 4 square patterned areas, composed of randomly placed triangular shapes. One of the 4 squares has a set of the triangular shapes placed in such a way that they create a circle, which can only be detected if the patient has some degree of stereopsis. The Frisby Stereotest began with familiarizing the child with 2-dimensional images of the patterns. The child was then tested with a “practice” plate (6 mm). The viewing distance for this plate varied from 20 to 40 cm. The plate was presented up to 4 times, rotating the location of the “circle” each time. If the child was able to detect the circular pattern on this plate 3 out of 4 times, he or she was tested using the 6 mm plate at a testing distance of 40 cm. If the child correctly identified the circular pattern on 3 out of 4 presentations with the 6 mm plate, the same procedure was repeated with a 3 mm plate and then a 1.5 mm plate. If the child was unable to detect the circular pattern on the 6 mm plate at 40 cm, he or she was tested at 30 cm. If the child was unable to detect the circular pattern on the 6 mm plate at 30 cm, the test was stopped.
The Randot Preschool test began with a pretest using nonstereo figures in Book 3 to familiarize the child with the shapes. After the pretest, polarizing glasses were placed on the child and the child was asked to identify the Randot shapes beginning with Book 3 (800 and 400 seconds of arc), followed by Book 1 (200 and 100 seconds of arc), and finally Book 2 (60 and 40 seconds of arc). If 2 shapes at a given level were identified correctly, testing proceeded to the next level. If the child could not correctly identify 2 shapes at a level, testing was stopped. Stereopsis was determined as the lowest level at which 2 shapes were correctly identified.
The Titmus Fly test was used to assess gross stereopsis. The child wore polarizing glasses for this test and the viewing distance was approximately 40 cm. The child was given time to adjust to the image and was then asked to touch the wings of the fly. If the child obviously pinched the wings above the test plate, credit was given for that level of stereopsis.
For each test, the examiner provided a confidence rating based on her assessment of the child’s behavior and cooperation with the testing. The 3-tiered confidence level was intended to specify the extent to which the examiner considered the results reliable: “very confident,” “somewhat confident,” or “not confident.”
Statistical Methods
The percentage of patients with stereopsis was compared between treatment groups using Fisher exact test. The median age at surgery and median visual acuity at age 4.5 years was compared between patients with and without stereopsis at age 4.5 years using the Wilcoxon rank-sum test. Multiple logistic regression was used to assess the combined effect of age at surgery and visual acuity at age 4.5 on the presence of stereopsis.
Results
Study Population
There were 114 children enrolled in the study, with 57 randomized to each treatment group. One patient in the intraocular lens group was lost to follow-up at age 18 months. The remaining 113 patients had visual acuity and stereopsis assessed at age 4.5 years (mean, 4.5 years; range, 4.5–4.9 years) and a clinical examination at age 5 years (mean, 5.0 years; range, 4.7–5.4 years) with an average length of follow-up of 4.8 years (range, 4.4–5.3 years) after cataract surgery. Significantly more secondary surgeries were required in the patients who received intraocular lenses compared with the patients who were left aphakic (41 vs 12 eyes).
Stereopsis
Of the 113 patients examined at age 4.5, 107 completed all 3 stereopsis tests and 6 patients were missing data for 1 or more of the tests: 3 patients were missing all the tests (phthisical eye, uncooperative child with poor vision, tester error); 2 patients were missing the Frisby and Randot Preschool tests (developmental delay, tester error); 1 patient was missing the Titmus test (afraid of the fly). Thus there were 110 patients with data for at least 1 of the 3 stereopsis tests.
Stereopsis for Each Test: Intraocular Lens vs Contact Lens Treatment Groups
For the Frisby test, 13 of 108 patients (12.0%) had some stereopsis (2 with 170 seconds of arc and 11 with 340 seconds of arc). The results did not differ significantly between the treatment groups (contact lens: 6 of 54 [11.1%], intraocular lens: 7 of 54 [13.0%], P = .99). For the Randot Preschool test, 4 of 108 patients (3.7%) had some stereopsis (2 with 400 seconds of arc and 2 with 800 seconds of arc). The percentages of patients with stereopsis were not significantly different between the treatment groups (contact lens: 3 of 54 [5.6%], intraocular lens: 1 of 54 [1.9%], P = .62). For the Titmus test, 21 of 109 patients (19.3%) had some stereopsis (3000 seconds of arc). The treatment groups were not significantly different (contact lens: 8 of 53 [15.1%], intraocular lens: 13 of 56 [23.2%], P = .34).
Among the 110 patients completing at least 1 of the tests, 28 patients (25.5%) had some stereopsis on 1 or more of the tests completed. Of these 28 patients, the number of patients showing positive stereopsis was as follows: all 3 tests: 3; Frisby and Titmus: 4; Frisby only: 6; Randot only: 1; Titmus only: 14. The percentages of patients with stereopsis on at least 1 of the tests were not significantly different between the treatment groups (contact lens: 11 of 54 [20.4%], intraocular lens: 17 of 56 [30.4%], P = .28).
Examiner Confidence Ratings
The percentage of patients for which the tester was very confident, somewhat confident, or not confident on the 3 stereopsis tests was as follows: Frisby (n = 107), very: 96 (90%), somewhat: 10 (9%), not: 1 (1%); Randot (n = 103), very: 99 (96%), somewhat: 3 (3%), not: 1 (1%); Titmus (n = 106), very: 99 (93%), somewhat: 5 (5%), not: 2 (2%).
Randot Stereopsis Responders
All 4 patients with Randot stereopsis results had acuity of 20/32 or better in the treated eye. Three of these patients (1 intraocular lens, 2 contact lens) had positive results on the other 2 stereopsis tests as well. All of these patients had surgery prior to 1.2 months of age. The examiner was very confident in the Frisby, Randot (400 seconds of arc), and Titmus responses in 2 patients, and somewhat confident in the Randot response (800 seconds of arc) in the third. The examiner was not confident in the fourth patient, who only had Randot stereopsis of 800 seconds of arc and no response to the other 2 tests.
Frisby Plus Titmus Responders
Four additional patients (2 intraocular lens and 2 contact lens) had positive responses to the Frisby and Titmus tests, but not to the Randot Preschool Stereoacuity test. Acuity in this group was also good (20/40 or better). Three of these patients had surgery prior to 1.1 months of age; the fourth patient had surgery at 6.0 months. The examiner was very confident in all of the results, both positive and negative, except for 1 “somewhat confident” on the Titmus Fly.
Frisby or Titmus Only Responders
Twenty patients demonstrated stereopsis on only 1 test (14 Titmus and 6 Frisby). Although many had acuities better than 20/40, 4 had acuity of 20/1333 or worse in the operated eye. Despite a vision that would preclude stereopsis on the test, the examiner was very confident of the results for 2 of these patients.
Best Stereopsis Responders
Two patients are of particular interest, because they demonstrated equal and excellent acuity and the highest level of stereopsis among those with positive stereopsis findings. Furthermore, the examiner’s rating was very confident for all of their stereopsis test results. These were the only patients to get 20/170 on the Frisby and 20/400 on the Randot Preschool Stereoacuity test. Both were contact lens patients ( Table 1 ).
| Age at Surgery (mo) | Visual Acuity at 4.5 Years | Stereopsis Tests (arc seconds) | |||
|---|---|---|---|---|---|
| Treated Eye | Fellow Eye | Frisby | Randot | Titmus | |
| 1.0 | 20/25 | 20/25 | 170 | 400 | 3000 |
| 1.1 | 20/12 | 20/12 | 170 | 400 | 3000 |
Stereopsis vs Age at Surgery and Visual Acuity at Age 4.5 Years
Age at surgery
The median age at surgery for patients with stereopsis on at least 1 of the tests was significantly younger than for patients without stereopsis (1.2 months vs 2.4 months, P = .002, Table 2 , Figure , Top panel). A significantly higher percentage of patients who had surgery before 49 days of age had some stereopsis compared to the patients who had surgery after 49 days of age (19 of 48 patients [40%] vs 9 of 62 patients [15%], P = .004).
| Characteristic | Stereopsis at Age 4.5 | P Value | |||
|---|---|---|---|---|---|
| Yes (N = 28) | No (N = 82 a ) | ||||
| Median | IQR | Median | IQR | ||
| Age at surgery (mo) | 1.2 | 1.0–2.0 | 2.4 | 1.2–3.4 | .002 |
| Visual acuity at 4.5 years, logMAR (Snellen) | 0.30 (20/40) | 0.20–0.75 (20/32–20/112) | 1.10 (20/252) | 0.50–1.70 (20/63–20/1002) | .0003 |
a For VA, the sample size for patients without stereopsis was 81 because VA could not be measured for 1 patient because of developmental delay.
