Strabological Findings after Macular Translocation Surgery with 360° Retinotomy

Abstract

purpose. To examine the strabological findings after macular translocation surgery with a 360° retinotomy.

methods. Thirty-two patients who underwent macular translocation surgery were divided into three groups based on their responses to the Bagolini striated lenses test: fusion, ignoring the image, and diplopia. The relevant factors affecting binocularity were compared among the three groups.

results. Five patients had peripheral fusion and three of these had gross stereopsis. Fifteen patients ignored the second image, and 12 patients had diplopia. The objective angle of macular rotation was smaller in the patients with peripheral fusion (15.0 ± 6.1°) than in those with diplopia (32.7 ± 11.7°). The subjective angle of cyclotorsion in those with peripheral fusion (6.0 ± 4.2°) was smaller than in those who ignored the image (“ignoring” group; 20.5 ± 9.19°) and the diplopia group (30.7 ± 12.8°). The amount of torsional sensory compensation in patients with diplopia (2.08 ± 3.83°) was significantly smaller than in those with peripheral fusion (9.00 ± 7.42°) and in the ignoring group (6.73 ± 3.86°). Patients with peripheral fusion were significantly younger (54.2 ± 14.3 years) than those in the ignoring group (67.7 ± 10.0 years) and those with diplopia (68.0 ± 5.4 years).

conclusions. Adaptive mechanisms are activated to reduce the surgically induced objective angle of cyclotorsion, and a cyclodeviation of 15° was the critical angle separating those who had peripheral fusion from those who did not. This value corresponds to the cyclofusional amplitude in normal adults.

Macular translocation surgery was first reported by Machemer and Steinhorst ¹ as an optional surgical treatment for age-related macular degeneration (ARMD) and has recently become a relatively common surgical procedure. Good results have been reported, ^{2 3} although the rotation of the macula results in an unavoidable cyclotorsion of the retina. ⁴ Patients with poor visual acuity in one eye usually adapt to the tilted image by a sensory reorientation of the spatial values of the retinal meridians, ⁵ but when patients have good visual acuity in both the surgically treated and untreated eyes, the diplopia can become intractable. 

There are several adaptive mechanisms used by subjects for the cyclodeviations such as cyclofusion, anomalous head posture, suppression, and anomalous retinal correspondence, ⁶ and the one selected varies. However, other visual abnormalities, resulting from the underlying macular lesion such as aniseikonia, central scotoma, and/or metamorphopsia, may disrupt the use of the adaptive mechanisms. ⁷ Thus, cyclotropia with image tilt and diplopia have been reported to be a significant complication of translocation surgery. 

To reduce the cyclotropia induced by macular translocation surgery, eyes have been rotated by strabismus surgery in combination with the translocation surgery. ^{8 9 10 11 12} However, the question arises on the degrees of cyclotorsion the patients can adapt without additional surgery. The purpose of this study was to examine the strabological findings in patients who underwent macular translocation surgery without concurrent strabismus surgery. 

Methods

All patients who underwent macular translocation surgery between January 1999 and December 2000 were referred to the strabismus clinic. The surgeries were performed at two institutions by two surgeons (HT, NO) and consisted of a 360° peripheral retinotomy followed by macular translocation for the treatment of subfoveal choroidal neovascularization. All the patients were examined by one of the authors (MS) for the evaluation of postoperative binocularity. Patients who had surgery in both eyes were not included. 

Thirty-two patients (18 men and 14 women; age range, 31–82 years, mean, 65.7) were studied. Twenty had ARMD, eight had high myopic chorioretinal degeneration, and four had polypoidal choroidal vasculopathy. There was no significant difference in the patients’ ages in these three groups. The patients’ characteristics are listed in Table 1 . 

The visual acuity in the surgically treated eye was worse than that in the untreated eye in all patients before surgery. The retina was rotated upward in 24 patients and downward in 8. The preoperative best corrected visual acuity (BCVA) ranged from hand motion to 20/30, and the postoperative BCVA at the most recent examination ranged from 20/500 to 20/25. The visual acuity was measured with the Standard Japanese Visual Acuity Chart and converted to the Snellen visual acuity or to the log of the minimum angle resolution (logMAR) for statistical purposes. The BCVA in the nonsurgical eyes varied from hand movements to 20/15. 

This research was conducted in accordance with the institutional guidelines of Nagoya University and conformed to the tenets of the World Medical Association Declaration of Helsinki. Informed consent was obtained from each patient for the surgery and the sensory and motor examinations. 

Surgical Methods for Macular Translocation

All macular translocation surgeries were performed with a modification of the 360° retinotomy method of Machemer and Steinhorst. ¹ Initially, a lensectomy with preservation of the anterior lens capsule was performed, followed by a complete removal of the vitreous by pars plana vitrectomy. Then four separate dome-shaped retinal detachments were created by a subretinal infusion of balanced salt solution. Fluid–air exchange led to the coalescence of the detachments and was followed by a 360° retinotomy at the ora serrata with automated scissors. This was followed by the injection of heavy perfluorocarbon liquid and the rotation of the whole retina around the axis of the optic disc with further injection of perfluorocarbon liquid during the rotation. The edges of the 360° retinotomy and the holes created for detaching the retina were sealed by endophotocoagulation. After this, an exchange of perfluorocarbon liquid by silicone oil (1000 centistoke) was performed. 

After 2 to 3 months, the silicone oil was removed in all cases. An IOL had been implanted in 2 eyes before the macular translocation surgery or was implanted in 18 eyes at the same time or after the silicone oil removal. The remaining 12 eyes were left aphakic because the visual acuity of the fellow eye was good; the patient wore a contact lens in the surgically treated eye and did not request an IOL implantation (five cases); the postoperative refractive error was small because of preoperative high myopia, and the patient wore glasses without difficulty, despite the anisometropia (three cases); the patient did not accept an IOL implantation or contact lens for fear of diplopia (three cases); or the fellow eye was also aphakic (one case). 

Postoperative complications, such as retinal detachment (one eye), proliferative vitreoretinopathy (one eye), and epiretinal membrane (one eye) were successfully treated, and these eyes were included. 

Sensory and Motor Examination

Sensory and motor examinations were first performed 3 to 6 months after the macular translocation surgery. The examinations were repeated at 3- and 6-month intervals, and the most recent results were used for the evaluation. The overall follow-up period was 4 to 23 months after the macular translocation surgery. Patients with aphakic eyes were examined with a disposable soft contact lens if anisometropia was greater than 6 D. 

Horizontal and vertical ocular alignments were measured by the alternate prism-and-cover test with the fixation target at 30 cm. Special care was taken to allow the patient sufficient time to refixate with the surgical eye during the prism-and-cover test. The degree of vertical deviation was recorded as a deviation of the surgically treated eye. Positive numbers represented hypertropia, and negative numbers represented hypotropia of the surgical eye. 

The subjective torsion was measured with the Maddox double-rod test. Red and white Maddox rods were placed in a trial frame with the red before the dominant eye and the white before the nondominant eye. The orientation of the glasses was aligned with the 90° marks of the trial frame. The patient then looked at a penlight through the Maddox rod lenses and was asked about the orientation of the red and white streaks. The examiner then turned the red and white rod until both lines were seen to be parallel. When a patient reported seeing two streaks but with the top of one slanted toward the nose, excyclotropia of that eye existed and was represented as a negative number. All patients reported seeing two streaks. 

For statistical analysis, the patients were divided into three groups according to their responses to the Bagolini striated lenses test: those with peripheral fusion, those who ignored the second image, and those who had diplopia. The Bagolini striated lenses were placed at 45° and 135°, and the patients fixated a target light at 30 cm. When the patients reported seeing a cross that intersected over the light source even if the brightness of the two streaks were different or a small central scotoma was noticed, they were classified as having peripheral fusion. When patients reported seeing two light spots or when they saw only one light but noticed two light streaks intersecting below or above the light, they were classified as having diplopia. Because all the patients saw two streaks with the Maddox rod test, when a patient reported seeing only one streak over the light, even if the patient noticed a dim light streak in the very periphery, he or she was classified as ignoring the second image. 

Patients used different mechanisms to deal with the diplopia, such as suppression, ignoring the second image, or the use of anomalous retinal correspondence. Some patients had an absolute scotoma due to the underlying disease and did not report a second image. It was very difficult to differentiate between these alternatives, and we therefore classified all patients who did not have diplopia or peripheral fusion into the ignoring category. 

Patients were asked whether they had diplopia under daily natural viewing conditions, and their answers were recorded separately from the responses to the Bagolini striated lenses test. Stereopsis was measured by a stereovision test (Titmus Stereotest; Oculus, Wetzlar, Germany). 

Measurement of Retinal Rotation

The objective angle of macular translocation was measured by comparing pre- and postoperative fundus photographs. The photographs were carefully taken with the patient’s head held as vertical as possible while the patient fixated a built-in fixation target. Patients who had a large central scotoma and were not able to see the target were asked to fixate the central area of the target. To determine the objective cyclotorsion, a line was drawn from the center of the optic nerve head to the junction of a retinal vessel near the fovea. The angle formed by this line and a horizontal line was measured on the pre- and postoperative fundus photographs. The magnitude of foveal rotation was calculated by subtracting the preoperative angle from the postoperative angle. A positive number indicated that the macula was rotated upward, and a negative number indicated downward rotation. For statistical purposes, absolute values were used as the amount of objective cyclotorsion. The difference between the objective and subjective cyclotorsion was designated as the sensory torsional compensation. 

Statistics

The data were analyzed by computer (StatView, ver.5; SAS Institute, Cary NC). To compare data from the three groups, one-way factorial analysis of variance (ANOVA) with the Scheffé test as a post hoc test was used. P < 0.05 was considered to be statistically significant. 

Because the patients with an IOL and those without may have experienced completely different visual conditions, this factor was carefully considered in the statistical analyses. We compared the patients’ age, pre- and postoperative BCVA, objective angle of retinal rotation, subjective angle of retinal rotation, and the amount of sensory torsional compensation between the two groups, with and without IOL implantation before starting the analysis based on the binocular viewing conditions. 

Results

There were no statistical differences between the patients with IOL and those without, and we combined the patients with and without IOL implantation based on the binocularity (Table 2) . The preoperative BCVAs and the BCVAs at the most recent examination are shown in Figure 1 . 

There was a high correlation between the subjective and objective angles of cyclotorsion (ρ = 0.975, P < 0.0001, Spearman rank correlation coefficient; Fig. 2 ). However, the absolute subjective cyclotorsion (22.0 ± 13.0°) was significantly smaller than that of the objective cyclotorsion (27.4 ± 11.8°; P < 0.0001, Wilcoxon signed rank test). There was a strong correlation between the amount of vertical deviation and the absolute objective cyclotorsion. (ρ = 0.733, P < 0.0001, Spearman rank correlation coefficient). All the patients had exotropia after surgery, and there was a significant negative correlation between the amount of horizontal deviation and the amount of absolute objective cyclotorsion (ρ = −0.653, P = 0.0003, Spearman rank correlation coefficient). 

Five patients (15.6%) had peripheral fusion with the Bagolini striated lenses test, and three of them passed the fly test (3000 seconds of arc; Titmus). Fifteen (46.9%) patients reported seeing only one streak and were classified in the ignoring group. Five patients in the ignoring group experienced double vision in their daily lives, and two of them underwent strabismus surgery. Four patients in the ignoring group reported diplopia when the vertical and horizontal deviations were neutralized by prisms. Twelve patients (37.5%) had diplopia, and seven of them received or were scheduled for strabismus surgery. The remaining five patients had diplopia, but they could manage it with or without optical correction and did not require any further treatment. Fresnel prisms were prescribed for one patient, and the patient was satisfied with them. 

Patients with peripheral fusion were significantly younger (54.2 ± 14.3 years) than patients in the ignoring group (67.7 ± 10.0 years) and patients with diplopia (68.0 ± 5.4 years; P = 0.031 and 0.034, respectively, one-way factorial ANOVA). The absolute objective angle of macular translocation was smaller in the patients with peripheral fusion (15.0 ± 6.12°) than in the patients with diplopia (32.7 ± 11.7°; P = 0.0132, one-way factorial ANOVA). The absolute subjective angles of cyclotorsion in patients with diplopia (30.7 ± 12.8°) and in patients in the ignoring group (20.5 ± 9.19°; P = 0.0004) were larger than in patients with peripheral fusion (6.0 ± 4.2°; P = 0.0358; one-way factorial ANOVA). The amount of torsional sensory compensation in patients with diplopia (2.08 ± 3.83°) was significantly smaller than in patients with peripheral fusion (9.00 ± 7.42°; P = 0.026) and in patients in the ignoring group (6.73 ± 3.86°; P = 0.042; one-way factorial ANOVA). The degree of horizontal deviation was larger in patients with diplopia (−22.9 ± 10.1 prism diopters [pd]) than in patients with peripheral fusion (−5.6 ± 1.8 pd; P = 0.0075, one-way factorial ANOVA). The vertical deviations in patients with diplopia (18.9 ± 7.7 pd) and in patients in the ignoring group (12.7 ± 7.0 pd) were larger than in patients with peripheral fusion (3.0 ± 2.6 pd; P < 0.001 and P = 0.036, respectively, one-way factorial ANOVA). 

The patients who had diplopia in daily life showed a larger amount of objective torsion (31.5 ± 12.3°) and subjective torsion (26.7 ± 11.8°) than those in patients with fusion (15.00 ± 6.12° and 6.00 ± 15.9°, P = 0.0474 and P = 0.0062, respectively, one-way factorial ANOVA) when cases with unilateral uncorrected aphakia were excluded. 

The correlations between the binocularity and the pre- and postoperative BCVAs, the BCVA of the fellow eye, and the postoperative follow-up period were not significant (one-way factorial ANOVA). 

Discussion

Despite the relatively large degree of retinal rotation, half of the patients did not report diplopia. It is interesting that five patients were able to obtain peripheral fusion as determined by the Bagolini striated lenses test, and three were able to appreciate gross stereopsis after the macular translocation without strabismus surgery. Unfortunately, there are not enough published data on the binocularity after macular translocation surgery because in many instances, simultaneous strabismus surgery was performed and observational study or statistical analysis was not conducted. ^{10 13} Furthermore, in none of these studies were the requirements for strabismus surgery mentioned. 

The macular translocation surgery was always performed on the worse eye. The BCVA of the fellow eye or the difference of the BCVA between the surgically treated eye and the fellow eye may be related to postoperative double vision. The patients in the fusion group had vision equal to or better than 20/200 in the worse eye, and the patients in the diplopia group had vision equal to or better than 20/250 in the worse eye. When the patients had BCVA equal to or worse than 20/500, they were placed in the ignore group. BCVA of 20/250 may be a critical vision for the perception of the image in the worse eye when both eyes are open. 

Torsional motor rotations up to approximately 8° have been demonstrated in normal subjects, ¹⁴ and normal adults are able to fuse approximately 8° of cyclodisparity by sensory fusion. ¹⁵ Combining sensory and motor cyclofusion, normal subjects have the potential of fusing up to 15° of cyclodisparity. ¹⁶ Also, some subjects have been reported able to fuse up to 30° of disparity after training. ¹⁷ It should also be remembered that patients who undergo macular translocation surgery do not have imbalances of the extraocular muscles as do those with paralytic strabismus. 

The average amount of macular translocation in the patients who obtained peripheral fusion was 15°, which corresponded notably to the 15° of cyclodisparity that can be fused. Vertical and horizontal deviations proportionally increased with the amount of macular rotation. All five patients who had peripheral fusion showed less than 10 pd of horizontal and vertical deviation. 

We found that three significant factors contributed to binocularity after macular translocation surgery: the patients’ age, the amount of macular rotation, and the amount of torsional sensory compensation. Younger patients had a better chance to attain fusion after macular translocation surgery, possibly because the older patients had undergone a greater amount of macular translocation (ρ = 0.396, P = 0.0293, Spearman rank correlation coefficient). Another possible reason is that younger patients have more potential to compensate for the diplopia, although this possibility was not supported by the results of statistical analysis. 

There may be other factors that influence postoperative binocularity. First, some of the patients who ignored the second image may have had preexisting horizontal deviations or suppression, possibly caused by the long-standing poor visual acuity of one or both eyes. Four of 15 patients who were classified into the ignoring group had a constant exotropia and suppression scotoma before surgery. Second, an absolute scotoma may have remained after the macular translocation surgery in the area of the previous macula. Thus, the image falling on the previous fovea would not be seen. In many instances after macular translocation surgery, patients did not switch to the surgical eye when the fellow eye was covered, as is usual in patients with strabismus. We placed prisms in front of the eye that ignored the second image to correct the vertical and horizontal deviations, and 4 of 15 patients then reported diplopia with the prisms. This finding suggests that when the image falls close to the new fovea, patients easily see the second image. 

Our results showed that some of the patients were able to perform daily visual tasks despite having diplopia. We suggest that in eyes at the first surgery, simultaneous strabismus surgery is not necessary if the subfoveal choroidal neovascular membrane is small and a large amount of retinal rotation is not necessary (<15°). When the lesion is large and more than 15° of macular rotation is planned, performing simultaneous or staged strabismus surgery is recommended to avoid intractable diplopia. 

In conclusion, attaining binocular vision after macular translocation surgery with a 360° retinotomy depends on the amount of macular rotation, the amount of torsional sensory compensation, and the patient’s age. Adaptive mechanisms are activated after the surgery to reduce the objective angle of cyclotorsion, and 15° was the critical angle for obtaining peripheral fusion, which corresponds to the cyclofusional amplitude in normal adults. 

 Supported by Grants-in-Aid for Scientific Research 12671702, 13307048, 14571665, 14370557, and 11470363 from the Ministry of Education, Culture, and Science of Japan.

 Submitted for publication February 19, 2002; revised September 4, November 12, and December 24, 2002; accepted January 15, 2003.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Patients’ Characteristics

Patient	Diagnosis	Eye	Age	Gender	IOL	Pre-op BCVA	Post-op BCVA	Fellow Eye BCVA	Dominant Eye	Objective Torsion	Subjective Torsion	Horizontal Deviation	Vertical Deviation	Bagolini	Daily Life Condition	Prism Correction	Comment	Follow-up Period (mo)
1	ARMD	R	50	M	+	20/100	20/133	20/13	L	−15	−5	−4	7RHT	Fusion	Fusion		Fly(+)	11
2	ARMD	R	61	M	+	20/500	20/200	20/40	L	30	30	−20	14LHT	Dip	Dip		OP	4
3	ARMD	R	62	M	−	CF	20/100	20/13	L	30	50	−25	30LHT	Dip	Dip		OP	15
4	ARMD	R	67	F	+	20/200	20/100	20/20	L	38	30	−18	20LHT	Dip	Int.Dip		OP	5
5	ARMD	R	69	M	+	20/600	20/60	20/13	L	35	35	−30	14LHT	Dip	Dip		OP	6
6	ARMD	R	69	M	+	20/100	20/50	20/20	L	25	20	−20	2LHT	Ignore	Dip	No Dip	Prism	10
7	ARMD	R	70	M	−	20/133	20/80	20/16	L	30	30	−10	14LHT	Ignore	No Dip	No Dip		12
8	ARMD	R	71	M	−	20/200	20/500	20/50	L	20	8	0	9LHT	Ignore	No Dip	No Dip		6
9	ARMD	R	72	F	−	20/500	20/200	20/30	L	25	20	−25	25LHT	Dip	Dip			6
10	ARMD	R	73	M	+	20/250	20/100	20/20	L	37	25	−16	6LHT	Ignore	No Dip	No Dip		15
11	ARMD	R	79	F	−	HM	20/133	HM	R	−23	−20	−8	12RHT	Ignore	No Dip	No Dip		12
12	ARMD	R	82	M	+	20/500	20/133	20/20	L	−40	−30	−18	10RHT	Ignore	No Dip	No Dip		9
13	ARMD	L	60	M	+	20/2000	20/200	20/13	R	15	10	−10	6RHT	Ignore	Int.Dip	No Dip		22
14	ARMD	L	62	F	+	20/250	20/100	20/13	R	15	10	−8	2RHT	Fusion	Fusion			23
15	ARMD	L	62	M	+	20/40	20/60	20/20	R	25	5	−5	4RHT	Fusion	Fusion			11
16	ARMD	L	63	M	+	20/100	20/50	20/16	R	−28	−30	−10	25LHT	Dip	Dip		OP	8
17	ARMD	L	67	M	−	20/200	20/250	20/16	R	45	40	−40	20RHT	Dip	Dip			5
18	ARMD	L	71	F	+	20/60	20/200	20/16	R	−20	−18	−18	12LHT	Dip	Dip			12
19	ARMD	L	71	F	−	20/200	20/133	20/30	R	22	25	−40	14RHT	Ignore	No Dip	Dip		11
20	ARMD	L	74	M	+	20/2000	20/133	20/20	R	40	25	−30	20RHT	Ignore	Int.Dip	No Dip		9
21	PCV	R	31	F	+	20/30	20/33	20/13	L	10	0	−4	2LHT	Fusion	Fusion		Fly(+)	11
22	PCV	R	77	M	+	20/300	20/40	20/16	L	−40	−50	−25	25RHT	Dip	Dip		OP	11
23	PCV	L	64	M	−	20/600	20/250	20/200	R	50	40	−20	20RHT	Ignore	Dip	No Dip	OP	5
24	PCV	L	66	M	+	20/200	20/200	20/13	R	10	10	−7	0	Fusion	Fusion		Fly(+)	8
25	Myopic	R	64	F	−	20/2000	20/200	20/100	L	41	30	−40	25LHT	Dip	Dip		OP	14
26	Myopic	R	66	F	+	20/200	20/40	20/133	L	10	5	−14	3LHT	Dip	Dip			11
27	Myopic	L	40	F	+	20/100	20/50	20/16	R	−30	−24	−18	18LHT	Ignore	Dip	No Dip	OP	12
28	Myopic	L	60	M	+	20/200	20/400	20/40	R	22	5	0	2RHT	Ignore	No Dip	No Dip		5
29	Myopic	L	63	F	−	20/200	20/60	20/25	R	18	20	−12	20RHT	Ignore	No Dip	Dip		13
30	Myopic	L	65	F	−	20/100	20/25	20/20	R	20	15	−14	12RHT	Ignore	No Dip	Dip		6
31	Myopic	L	75	F	+	20/1000	20/100	20/40	R	30	20	−10	25RHT	Ignore	No Dip	Dip		12
32	Myopic	L	77	F	−	20/600	20/60	20/33	R	−30	−30	−10	14LHT	Dip	Dip			8

Table 2.

Figure 1.

 

Pre- and postoperative BCVA.

Figure 2.

 

Relationship between objective cyclotorsion and subjective cyclotorsion. Solid line: y = 1.367 + 0.845 · x; R ² = 0.945. Dashed line: y = x.