Japanese Journal of Ophthalmology

, Volume 55, Issue 4, pp 383–388

Influence of timing of initial surgery for infantile esotropia on the severity of dissociated vertical deviation

Authors

    • Yagasaki Eye Clinic
    • Department of OphthalmologySocial Insurance Chukyo Hospital
  • Yoshimi O. Yokoyama
    • Department of OphthalmologySocial Insurance Chukyo Hospital
  • Mariko Maeda
    • Department of OphthalmologySocial Insurance Chukyo Hospital
Clinical Investigation

DOI: 10.1007/s10384-011-0043-1

Cite this article as:
Yagasaki, T., Yokoyama, Y.O. & Maeda, M. Jpn J Ophthalmol (2011) 55: 383. doi:10.1007/s10384-011-0043-1

Abstract

Purpose

The purpose of this study was to clarify whether the timing of initial surgery for infantile esotropia contributes to better sensory outcomes and to the severity of dissociated vertical deviation (DVD).

Methods

This retrospective study examined 55 children, aged 8 years and younger, who underwent esotropia surgery. Based on age at the time of the surgery, patients were divided into the very early surgery group (birth to 8 months, N = 14), early surgery group (9–24 months, N = 23) and late surgery group (25 months and older, N = 18). Sensory and motor outcomes of the three groups were statistically evaluated.

Results

No significant differences in the incidences of DVD were noted among the three groups at the final visit. However, all DVDs of the very early surgery group were latent, whereas 38.9% of the DVDs for both near and distance of the late surgery group were manifest. Significant differences were found among the three groups for the proportion of patients with manifest DVD, latent DVD and without DVD, and for the incidence of additional surgery for manifest DVD (p < 0.05, G-test).

Conclusion

Early surgery for infantile esotropia decreases the severity of DVD and lowers the need for an additional operation for DVD.

Keywords

Dissociated vertical deviationInfantile esotropiaFusionTiming of surgery

Introduction

Dissociated vertical deviation (DVD) is a spontaneous or dissociation-induced elevation of the non-fixating eye. This is often accompanied by abduction and extortion, and is usually bilateral and frequently asymmetric. DVD is often associated with a disruption of binocular vision early in life, which may be caused by deprivation amblyopia due to congenital cataracts or corneal dystrophy, or by any type of strabismus, especially infantile or congenital esotropia [1, 2]. The mechanism and nature of DVD are not well understood [3, 4], and it is reported that DVD can be observed in subjects with normal binocular vision [5].

Infantile esotropia is defined as a large constant angle esotropia that is confirmed by 6 months of age. In addition, multiple clinical motor or sensory abnormalities are also reported to be associated with infantile esotropia. These motor characteristics include DVD, monocular smooth pursuit asymmetry, latent nystagmus and inferior oblique overaction. The etiology of this esotropia may be based on either the Chavasse [7] or the Worth [6] theory. The Chavasse theory has led to progressively earlier surgical intervention for infantile esotropia, as this is thought to result in better sensory outcomes [815], especially with surgery prior to 6 months of age or within the duration of the misalignment for 6 months after the onset of esotropia [14, 16].

The relationship between the earlier timing of the alignment of esotropia and the change of motor abnormalities remains unclear [1618]. On the other hand, it is well known that most esotropic patients with DVD do not require specific therapy, as peripheral binocular fusion will eventually control the DVD once the horizontal deviation is corrected. However, no specific optical or surgical treatment exists that can completely eliminate DVD.

The purpose of this study was to clarify the influence of the timing of the initial surgery for infantile esotropia on the sensory outcomes for the severity of DVD.

Methods

Between January 1997 and December 2009, we performed a retrospective review of the medical records of 71 consecutive patients who were all under 8 years of age and had undergone strabismus surgery for infantile esotropia. The infantile esotropia inclusion criteria were (1) onset of constant esotropia prior to 6 months of age and (2) comitant esotropia of 20 or more prism diopters (PD) with full cycloplegic correction.

At the initial visit, all patients underwent a routine, comprehensive evaluation that included cycloplegic refraction with 1% atropine sulfate eyedrops. Patients with hyperopia greater than 1.5 diopters (D) or astigmatism greater than 1.5 D were prescribed a pair of full corrective glasses to minimize the accommodative effects on ocular deviation. Patients with hyperopia greater than 3.0 D who showed a decrease in esodeviation of less than 10 PD after spectacle correction were included in this study.

Horizontal deviation was measured using the alternate prism and cover test (APCT) on non-accommodative targets for both distance and near to minimize the accommodative stimulus, with patients wearing their glasses when necessary. In order to avoid inaccurate measurements, neither the Krimsky nor the Hirschberg methods were applied [19]. The prism under cover test was used to measure DVDs as described elsewhere [20]. DVDs were categorized as being latent when slow elevation with only one eye occluded and brisk recovery toward the primary position after removal of the occluder were noted, and as manifest when slow elevation of either eye was found to occur spontaneously or when no moving downward of drifting eye after removal of the occluder was found. For children too young to undergo either the APCT or the prism undercover test at distance, accurate measurements were obtained for near only. All deviations were repeatedly measured on at least three occasions before surgery, and all maximum angles of deviation were applied in this study. During the follow-up period, the measurement and the categorization of DVD were performed with a prism neutralization of horizontal deviation by one of the authors.

Visual acuity was measured by a variety of methods appropriate for age. Grating acuity in preverbal children was assessed using either Teller Acuity Cards (Vistech Consultants Inc, Dayton, OH) or Cardiff Acuity Cards (Keeler, Windsor, UK). As patients became verbal, best-corrected visual acuity (BCVA) was obtained using the single Landolt ring chart. Amblyopia was defined as a difference in grating acuity of one or more octaves or in BCVA of two or more lines between both eyes.

The fusion status was assessed using the Bagolini striated lenses test at both distance and near. The response was defined as fusion when two crossed streaks either with or without a central scotoma were seen, or as suppression when the streaks were never seen at the same time. Stereo acuity was measured using the Titmus stereotest (Stereo Optical, Chicago, IL), the Randot stereotest (Stereo Optical) and the Randot Preschool Stereoacuity Test (Stereo Optical).

All patients underwent a bimedial rectus recession procedure (BMR) measured from the limbus based on the surgical protocol by Helveston [2]. To enable a unified consideration of the influence of horizontal surgery on DVD, five patients undergoing either simultaneous surgery for DVD or inferior oblique overaction at the time of the initial rectus recession were excluded from this study, and the remaining patients received only BMR as the first procedure. Patients were also excluded if they had previous strabismic surgery, any history of developmental delays, preterm birth with less than 37 weeks’ gestational age, Down syndrome, or any neurological or ocular abnormalities other than esotropia. A total of 55 patients met all criteria and were analyzed for the study. The minimum postoperative follow-up was 9 months. Postoperative follow-ups were performed at 1, 2, 4 and 8 weeks. Subsequently, patients were then seen every 3–6 months, on average.

To determine the influence of the timing of the initial surgery for infantile esotropia on the severity of DVD, the 55 patients who met the inclusion criteria for this study were divided into three subgroups based on the age at which they initially underwent the esotropia surgery. Fawcett et al. [21] report that the critical period for susceptibility of stereopsis begins at 2.4 months and then peaks at 4.3 months in children with infantile esotropia. Therefore, as the duration of misalignment was 6 months from the onset of stereopsis, patients undergoing alignment by 8 months of age were defined as the very early surgery group. The early and late surgery groups were classified as alignments between the ages of 9–24 months and those that occurred at 25 months or later, respectively.

After entering the collected data into Microsoft Excel (Microsoft Inc., Redmond, WA), the statistical analyses of the data were performed with StatMate III for Macintosh (ATMS Co., Tokyo, Japan) with the help of a biostatistician when needed. Subgroup means were compared using a two-tailed t-test, with comparisons obtained through the use of a G-test or Z-test for differences in proportions. The study was approved by the Clinical Research Ethics Committee of Yagasaki Eye Clinic and the Social Insurance Chukyo Hospital, and conducted in accordance with the tenets of the Declaration of Helsinki.

Results

The number of patients in the very early, the early and the late surgery groups were 14, 23 and 18, respectively. Although there were statistically significant differences among the three groups for the age at first visit and the age at surgery, there were no significant differences among the three groups for the age at the final visit and the duration of follow-up after surgery (Table 1). The preoperative angle of deviation in the late surgery group was significantly smaller than those seen for both the very early and the early surgery groups. Of the 55 patients, 52 (94.5%) were initially aligned to within 10 PD of orthotropia. The remaining two undercorrected patients of the early surgery group and one overcorrected patient of the late surgery group were all heterophoric at both near and distance. At the final visit, although 2 patients in the very early surgery group, 5 patients in the early surgery group and 6 patients in the late surgery group were not within 10 PD of orthotropia at distance, only 2 patients in the early surgery group and 2 patients in the late surgery group were exotropic, and the others showed intermittent exotropia or esotropia. The angle at 1 week after surgery and at the final visit did not differ significantly for the three groups (Table 1). At the final visit, all groups showed a small exo-shift of the angle of deviation for near vision at 1 week after surgery (mean of 2.00, 4.00 and 7.11 PD). However, these changes of deviation were not statistically significant (p > 0.05, Aspin-Welch t-test).
Table 1

Surgical data and ocular deviations of patients

 

Very early surgery

Early surgery

Late surgery

p value

(≤8 months)

(8–24 months)

(>24 months)

Number

14

23

18

 

Age at first visit (years)

0.42 ± 0.12

1.62 ± 1.70

3.44 ± 2.54

p < 0.001*

Age at surgery (years)

0.53 ± 0.09

1.29 ± 0.43

4.30 ± 1.70

p < 0.001*

Age at final visit (years)

8.60 ± 3.29

8.80 ± 3.67

11.07 ± 3.58

N.S.**

Follow-up (years)

8.07 ± 3.67

7.51 ± 3.71

6.77 ± 3.78

N.S.**

Angle of deviation (PD)

 At initial visit

45.36 ± 8.43

46.74 ± 12.02

34.17 ± 8.45

p < 0.001**

 At 1 week postop

1.14 ± 4.55

2.70 ± 5.76

−0.61 ± 5.85

N.S.**

 At final visit (N)

−0.86 ± 10.73

−1.30 ± 9.63

−7.72 ± 16.19

N.S.**

 At final visit (D)

−1.43 ± 8.19

−0.83 ± 9.52

−5.94 ± 13.12

N.S.**

PD Prism diopter, N at near, D at distance

* Aspin-Welch t-test, ** Student's t-test

At the first visit, DVD was detected in 0.0% of the very early group, 17.4% of the early group and 27.8% of the late surgery group. The relationship between the incidence of DVD and the timing of the first visit were statistically significant (Table 2, p < 0.05, G-test). At the final visit, DVD at near was present in 57.1% of the very early, 52.2% of the early and 66.7% of the late surgery groups. DVD at distance was also present in 57.1% of the very early, 56.5% of the early and 61.1% of the late surgery groups. No significant differences were found among the three groups in the incidence of DVD at both near and distance (Table 2, p > 0.05, Z-test). However, all DVDs of the very early surgery group were latent at both near and distance, whereas more than half of the DVDs of the late surgery group were manifest. In the degree of the severity of DVD, there was a statistically significant difference in the proportion of patients with manifest DVD, latent DVD and without DVD in the three groups (Table 2, p < 0.05, G-test).
Table 2

DVD at initial and final visits

 

Very early surgery

Early surgery

Late surgery

p value

(≤8 months)

(8–24 months)

(>24 months)

At initial visit

0 (0.0%)

4 (17.4%)

5 (27.8%)

p < 0.05*

At final visit

 At near

8 (57.1%)

12 (52.2%)

12 (66.7%)

N.S.**

  Bilateral

(2)

(7)

(1)

 

  Unilateral

(6)

(5)

(11)

 

  Size (PD)

5.40 ± 1.84

6.21 ± 3.49

6.42 ± 3.73

N.S.***

  Manifest DVD

0 (0.0%)

2 (8.7%)

7 (38.9%)

p < 0.05*

  Latent DVD

8 (57.1%)

10 (43.5%)

5 (27.8%)

 

  DVD (−)

6 (42.9%)

11 (47.8%)

6 (33.3%)

 

 At distance

8 (57.1%)

13 (56.5%)

11 (61.1%)

N.S.**

  Bilateral

(3)

(7)

(1)

 

  Unilateral

(5)

(6)

(10)

 

  Size (PD)

5.38 ± 1.85

5.58 ± 3.78

7.00 ± 3.36

N.S.***

  Manifest DVD

0 (0.0%)

2 (8.7%)

7 (38.9%)

p < 0.05*

  Latent DVD

8 (57.1%)

11 (47.8%)

4 (22.2%)

 

  DVD (−)

6 (42.9%)

10 (43.5%)

7 (38.9%)

 

PD Prism diopter

* G-test, ** Z-test, *** Student's t-test

At the initial visit, strabismic amblyopia was found in 1 patient of the very early, 2 patients of the early and 2 patients of the late surgery groups. During the follow-up period, an additional 3 patients of the early and 2 patients of the late surgery groups showed either anisometropic or secondary microtropic amblyopia. All patients with amblyopia were treated with occlusion therapy consisting of 2–4 h part-time occlusion using either adhesive patches or eyeglass patches and/or fully corrected glasses. The relationship between the incidence of amblyopia and the timing of surgery was not statistically significant (p > 0.05, G-test). Furthermore, only one of the amblyopic patients showed manifest DVD, and another 5 patients did not reveal any DVD. These findings suggest that the severity of DVD was not associated with either amblyopia or occlusion therapy.

The results of the sensory assessment for fusion and stereopsis at the final visit were obtained from 51 and 49 patients, respectively (Table 3). Patients who were too young to undergo the sensory testing reliably were excluded. The percentage of patients who achieved fusion at near in the very early, the early and the late surgery groups were 83.3, 66.7 and 38.9%, respectively. The timing of the alignment significantly affected the prevalence of fusion at near (Table 3, p < 0.05, G-test), although no significance was noted for the relationship between the surgical timing and the prevalence of fusion at distance. The timing of alignment also significantly affected the prevalence of stereopsis after treatment, ranging from 50.0% in the very early surgery group down to only 11.1% in the late surgery group (Table 3, p < 0.05, G-test). Although only 1 out of 9 patients (11.1%) with manifest DVD demonstrated fusion at both near and distance, 12 out of 19 patients (63.2%) without DVD and 13 out of 23 patients (56.5%) with latent DVD demonstrated fusion at both near and distance. The relationship between the severity of DVD and the achievement of fusion was statistically significant (Table 4, p < 0.05, G-test).
Table 3

Sensory outcome at final visit

 

Very early surgery (≤8 months)

Early surgery (8–24 months)

Late surgery (>24 months)

p value

Fusion (number)

12

21

18

 

 At near

10 (83.3%)

14 (66.7%)

7 (38.9%)

p < 0.05*

 At distance

7 (58.3%)

9 (42.9%)

4 (22.2%)

p = 0.1303*

Stereopsis (N)

12

19

18

 

 Stereopsis (+)

6 (50.0%)

3 (15.8%)

2 (11.1%)

p < 0.05*

  200” or better

2 (16.7%)

1 (5.3%)

0 (00.0%)

 

  3000” or better

4 (33.3%)

2 (10.5%)

2 (11.1%)

 

 Stereopsis (−)

6 (50.0%)

16 (84.2%)

16 (88.9%)

 

G-test

Table 4

Relationship between the severity of DVD and fusion

 

Fusion N (+) and D (+)

Fusion N (+) and D (−)

Fusion N (−) and D (−)

DVD (−)

12

4

3

Latent DVD

13

6

4

Manifest DVD

1

2

6

N Near, D distance

p < 0.05, G-test

During the follow-up period, although none of the very early group required additional surgery for DVD, surgery for DVD was performed for two (8.7%) of the early group and 5 (27.8%) of the late group. The relationship between the surgical timing for esotropia and the additional surgery for DVD was statistically significant (Table 5, p < 0.05, G-test).
Table 5

The relationship between the surgical timing for esotropia and the additional surgery for DVD

 

Very early surgery (≤8 months)

Early surgery (8–24 months)

Late surgery (<24 months)

Surgery for DVD (+)

0

2

5

Surgery for DVD (−)

14

21

13

p < 0.05, G-test

Discussion

The Worth [6] theory states that esotropia results from a primary motor fusion defect with a secondary sensory fusion deficiency, whereas the Chavasse [7] theory hypothesizes that esotropia is caused by a primary motor dysfunction with a secondary loss of binocularity due to abnormal visual stimulation even though patients with infantile esotropia do have a potential cortical fusion ability at birth. However, based on this theory, strabismus surgeons have been performing surgical interventions for infantile esotropia at progressively earlier ages. After the first report by Wright et al. in 1994, it was established that an earlier surgical timing for infantile esotropia by 6 months of age could result in a better sensory outcome, especially for finer stereopsis [814].

The question is whether this sensory trend could provide a better motor outcome as well? Wright [22] states in his textbook on the history of DVD that children who undergo early surgery for infantile esotropia to improve their binocularity may have less DVD. He also points out that during the 1970s and early 1980s, surgery was commonplace for large, cosmetically obvious DVD. In those children that underwent early intervention for infantile esotropia, the presence of DVD was minimal and could only be demonstrated by prolonged cover testing. When compared to reports from 30 years ago, the incidence of DVD in children who had early surgery for infantile esotropia was found to remain unchanged. However, with the advent of early intervention for infantile esotropia, it was found that these children could achieve better binocular fusion. While this ultimately reduced the severity, it did not decrease the incidence of DVD, although the number of patients that required DVD surgery did decrease. Recent reports on the incidence of DVD in patients who have undergone early surgery for infantile esotropia remain unchanged compared to earlier reports (64–92%) [1618]. Arslan et al. [23] report that an incidence of DVD of 24.1% in patients surgically treated between ages of 6 months and 2 years was statistically smaller than 52.0% in patients treated after age 2 years, but their mean follow-up period of 2.6 years was not long enough to estimate the incidence of DVD after esotropia surgery because of the mean age at onset of DVD of 2.8 years reported by Neely et al. [18]. In the present study we also found an incidence of DVD for the very early surgery group of 57.1%, similar to previous reports [1618]. However, the one difference in our study is that we were able to prove that the earlier timing of the initial surgery for infantile esotropia had a statistically significant influence on the severity of DVD.

Thus, the question is what mechanism is responsible for reducing the severity of DVD in patients who undergo an earlier surgery for infantile esotropia? Since the first report of DVD in 1895 [24], the etiology of DVD has remained elusive. Guthrie and Wright [25] hypothesize that the development of DVD is the clinical result of a lack of inhibitor input from the binocular cells in the occipital cortex to the rostral interstitial nucleus of the medial longitudinal fasciculus as the vertical gaze center. Guyton [3] speculates that the vertical vergence component of DVD might serve to dampen the horizontal component of latent nystagmus, which then causes a decrease in vision. Brodsky [4] also proposes that the dorsal light reflex as a primitive visuovestibular reflex function could manifest as DVD in humans when early onset strabismus precludes the development of normal binocular vision. Van Rijn et al. report an either asymmetrical or symmetrical vertical phoria simulating DVD in ophthalmologically normal subjects. They suggest that the DVD that is often observed in subjects with a disruption of binocular vision early in life reflects the enhancement of a phenomenon that is present in normal subjects as well [5]. Wang and Bedell [26] also confirm these results in normal subjects exhibiting little or no vestiges of the dorsal light reflex. When the reports are taken together with our present results, DVD appears to originate from a normally observed vertical phorial movement. This can be controlled so that it is very small and strongly latent when there are well-developed binocular conditions. The severity of the DVD is significantly associated with fusion, which can be strongly established when there is an earlier timing of surgery for infantile esotropia.

It is well known that DVD is a spontaneous drifting of either eye upward when the patient is fatigued or daydreaming, or when fusion is artificially interrupted by covering one eye. Parents often report the eye as going way up and out at home, especially when their child is daydreaming. The parents’ description seems helpful to estimate the condition of DVD. However, DVDs are associated with infantile esotropia most commonly and often found in other forms of horizontal deviation, especially intermittent exotropia. If the patient has fatigue, is daydreaming or experiences artificial interruption of fusion, it causes complicated horizontal deviation leading to heterotropia and then to heterophoria after which latent DVD can be altered to become artificially manifest. We believe that the correct judgment of the severity of DVD should be performed without any influences of horizontal deviations, and most of the recent investigators estimate both the quality and quantity of DVD after the alignment of horizontal deviation in the same fashion.

The present study may be subject to limitations, as our sample size was small, and this study was a retrospective analysis. At all examinations, the classification based on the timing of surgery was not masked to us, and this may have influenced the results. However, this report is the first time that the relationship between the timing of surgery for infantile esotropia and the severity of DVD has been specifically examined. The increasingly early intervention for infantile esotropia now plays an important role in helping to ensure there is a better sensory outcome, including fusion. Based on the relationship we determined between the severity of DVD and fusion, earlier surgery for infantile esotropia may potentially decrease the risk of having to undergo additional operations for DVD, although the incidence of DVD still remains high. Further studies are needed to elucidate the factors that contribute to better motor outcome in patients with infantile esotropia.

Acknowledgments

An earlier version of this paper was presented at the 66th Annual Meeting of the Japanese Association of Strabismus and Amblyopia, 2–3 July 2010, Tokyo, Japan. The authors would like to thank all of the orthoptists at the Yagasaki Eye Clinic, and at Social Insurance Chukyo Hospital’s Department of Ophthalmology for their assistance in preparing this manuscript.

Copyright information

© Japanese Ophthalmological Society 2011