- Email: [email protected]
Assessment of Vision in Idiopathic Macular Holes with Macular Microperimetry Using the Scanning Laser Ophthalmoscope
Assessment of Vision in Idiopathic Macular Holes with Macular Microperimetry Using the Scanning Laser Ophthalmoscope Raymond N. Sjaarda, MD, Deborah A. Frank, BS, Bert M. Glaser, MD, John T. Thompson, MD, Robert P. Murphy, MD Background: Visual loss in eyes with full-thickness macular holes has been thought to be due to the absence of retinal function in the area of neurosensory defect as well as loss or reduction of retinal function in the surrounding area of neurosensory retinal detachment. With the advent of surgical techniques to treat macular holes, it is increasingly important to better characterize this visual dysfunction. Methods: Thirty eyes of 30 patients with full-thickness idiopathic macular holes were evaluated with microperimetry using the scanning laser ophthalmoscope to detect and quantitate absolute and relative scotomata within the central 40° of visual field. A log 2 scale of test stimulus intensities was established. Results of microperimetry were compared with best-corrected visual acuities as measured on the logarithmic Early Treatment of Diabetic Retinopathy Study chart as well as duration of symptoms. Results: All 30 eyes showed an absolute scotoma in the area of neurosensory defect as well as surrounding relative scotomata in the area of neurosensory detachment. Best-corrected visual acuity was correlated with the size of the absolute and relative scotomata (P < 0.002). The sizes of the scotomata were correlated with the duration of symptoms of the macular holes (P < 0.05). Conclusion: Microperimetry using the scanning laser ophthalmoscope demonstrates that the visual loss associated with macular holes is related to the reduction of retinal function in the area of the surrounding neurosensory detachment as well as the absence of retinal function in the area of neurosensory defect. The size of the scotomata, determined by microperimetry, is correlated with the patient's visual acuity as well as the duration of symptoms of the macular hole. Ophthalmology 1993;100:1513-1518
With the advent of promising new vitreous surgical techniques to treat and restore vision in eyes with macular
Originally received: December 22, 1992. Revision accepted: March 8, 1992. From the Retina Institute of Maryland, Baltimore, Maryland. Presented in part at the Annual Meeting of the Vitreous Society, Laguna Niguel, October 1992. Reprint requests to Raymond N. Sjaarda, MD, Retina Institute of Maryland, O'Dea Medical Arts Bldg, 7505 Osler Dr, Suite 103, Baltimore, MD 21204.
holes 1,2 there has been renewed interest in better characterizing precisely the associated visual alterations. Visual loss associated with macular holes is thought to be related to the neurosensory defect as well as the surrounding neurosensory detachment. 3- s However, previous reports have had limited success in demonstrating absolute and/or relative scotomata corresponding to the macular holes,s-7 and have not shown any correlation of these to visual acuity. We present our findings of absolute and relative scotomata as demonstrated by microperimetry using the scanning laser ophthalmoscope (SLO) in a consecutive series of 30 patients with idiopathic macular holes.
1513
Ophthalmology
Volume 100, Number 10, October 1993
Patients and Methods A total of 30 eyes of 30 consecutive patients with idiopathic macular holes and symptoms of reduced vision were studied. Only eyes with 1+ nuclear sclerosis or less were included in this study group. Best-corrected visual acuity was obtained by protocol refraction and measurement by a trained clinical coordinator using the logarithmic eye chart used in the Early Treatment of Diabetic Retinopathy Study.8 Visual acuities were converted to visual acuity scores for statistical comparison (see below). 8 All patients underwent examination, including slitlamp biomicroscopic, funduscopic, and fundus biomicroscopic examination with contact lens or 78-diopter (D) indirect condensing lens. Fundus photography and fluorescein angiography were performed. The diagnosis of a macular hole was made by the presence of a full-thickness neurosensory defect and the presence of a window defect on fluorescein angiography. Holes were staged according to the classification proposed by Gass. 3.5 Microperimetry was performed on all patients using the Rodenstock SLO (Rodenstock, USA, Danbury, CT) and a software package developed for microperimetry using this instrument (Scanning Laser Ophthalmoscope Psychophysics Programs, Eye Research Institute of Retina Foundation, Boston, MA). All microperimetry was performed by two examiners (RNS and OAF). The SLO unit had helium-neon, argon blue-green, and infrared lasers. To provide for optimal imaging as well as mesopic background illumination, the helium-neon laser was adjusted to an output of 1 to 2 poW and the infrared laser was adjusted for an output of 20 poW. The argon laser was not used. For all tests, 40° fields were used. Fixation intensity as well as isopters of stimulus intensity were established in the following manner. Contrast of fixation and test stimuli can be adjusted by the software from 0% to 100% in 256 increments (0 through 255) based on an 8-bit computer microprocessor (2 8 = 256) producing "shades of gray." For fixation, a 100% contrast (255) cross with a size of 10 X 10 to 30 X 30 pixels was used. For a test stimulus, a 6 X6-pixel square (equivalent to 19.7 minutes of arc square, corresponding to approximately 97 porn square on the retina) was used for kinetic testing. A total of five contrast settings were used for the test stimulus to create five isopters for kinetic testing. The five contrast settings of 15,31,63,127, and 255 vary by log 2 contrast intensity (2n - 1), thus creating logarithmic increments of intensity. Kinetic perimetry was performed similar to that of testing with a Goldmann kinetic perimeter. The examiner is able to view the macular pathology and assess fixation in "real time" while performing the micro perimetry. Patients were asked to fixate on the fixation cross and signal when the test stimulus was first seen. lsopters were mapped on an acetate overlay on the video monitor and subsequently mapped out using the perimetry software. The software provides digital analysis which converts the area mapped into pixels. Statistical comparisons were performed using multivariate analysis (for vision and scotoma comparisons) or linear regression (for scotoma and duration of symptom comparisons).
1514
Results
------------------------------There were 20 women and 10 men in the study. The
average age was 64.6 years (range, 53-75 years). The average best-corrected vision was 20/80-3 (range, 20/4020/200+2). All patients noted symptoms ofloss of vision and/or metamorphopsia. The average duration of symptoms before presentation was 33.7 weeks (range, 1-100 weeks). There were 5 stage 2 holes, 18 stage 3 holes, and 7 stage 4 holes. Absolute and relative scotomata were demonstrated in all patients. All patients showed an absolute scotoma corresponding exactly to the size of the neurosensory defect. Retina at the rim of the macular hole perceived the brightest (255) intensity stimulus in all eyes. The 127intensity stimulus was perceived in 12 of 30 eyes by the retina at the margin of the hole. In the remaining 18 patients, the 127-intensity isopter was seen further away from the margin of the hole, creating a surrounding isopter of relative scotoma. All patients showed additional surrounding scotomata of progressively dimmer isopters (Figs 1 and 2). These scotomata were generally within the area of visible neurosensory detachment. However, the two dimmest isopters, the 31- and 15-intensity levels, often were located outside the area of apparent neurosensory detachment. Vision was a function of both the absolute scotoma and the surrounding relative scotoma (Fig 3). Best-corrected visual acuity was correlated to the size of the absolute and the 15-intensity isopter relative scotoma by multivariate analysis using the two as independent variables (P < 0.002). The average radius of the absolute scotoma was 13.5 pixels (44.3 minutes of arc, or 219 porn on the retina). The minimum and maximum radii were 5.5 pixels (18.0 minutes of are, or 89 porn on the retina) and 25.4 pixels (83.3 minutes of are, or 412 porn on the retina), respectively (Fig 4). The average, minimum, and maximum radii of the 15-intensity relative scotoma were 70.4 pixels (230.9 minutes of are, or 1142 porn on the retina), 37.3 pixels (122.3 minutes of are, or 605 porn on the retina), and 112.2 pixels (368.0 minutes of arc, or 1820 porn on the retina), respectively (Fig 4). Figure 5A shows a comparison of the size of absolute scotoma (which equals the size of the neurosensory defect) and duration of symptoms. Eyes with a longer duration of symptoms, and presumably older holes, had significantly larger holes (R 2 = 0.40; P < 0.05). Figure 5B shows a comparison of the size of the 15-intensity isopter relative scotoma with the duration of the patients' symptoms. Eyes with a longer duration of symptoms showed significantly larger radii of the 15 isopter scotoma (R2 = 0.37; P < 0.05).
Discussion Using kinetic microperimetry with the SLO, we demonstrated absolute and relative scotomata in 30 eyes of 30 consecutive patients with idiopathic macular holes. In addition, we validated the logical hypothesis that visual loss in this condition is related to the absence of retinal func-
Sjaarda et al . Idiopathic Macular Holes
Figure 1. A 63-year-old woman with a 44-week history of poor vision in her left eye is shown. Best-corrected visual acuity was 20/80-2. A, red-free photograph shows a full-thickness stage 3 macular hole and the surrounding rim of a neurosensory detachment. B, fluorescein angiogram demonstrates a window defect in the area of a neurosensory detachment. C, scanning laser ophthalmoscopic photograph shows neurosensory defect and surrounding area of darker discoloration, possibly delimiting the full extent of the neurosensory detachment. D, macular microperimetry shows the intensity isopters of scotomata of the macular hole and surrounding retina; from inner to outer, they are 255 (absolute), 127, 63, 31, and 15.
tion in the area of neurosensory defect as well as the reduction of retinal function in the area of the surrounding neurosensory detachment. Previously described techniques using an Amsler grid 5,6 or hybrid forms of microperimetry4 have demonstrated absolute and/or relative scotomata with limited consistency. Johnson and Gass5 demonstrated a scotoma, or a scotoma with metamorphopsia, on Amsler grid testing in as many as 40% of 109 patients with stage 3 or 4 macular holes. Smith et al,6 using Amsler grid techniques, demonstrated an absolute scotoma in 8 of 24 eyes of patients with full-thickness macular holes. They calculated the scotoma to be less than 30 in diameter in these patients. No absolute scotoma could be perceived in 16 of24 eyes, but all of these eyes demonstrated metamorphopsia. They concluded that the predicted central scotoma would be less than lOin diameter « 1 square on an Amsler grid) and therefore would not be recognized. They also concluded that accurate plotting of small central scotomata is problematic because of the difficulty in maintaining
fixation. Weare in agreement that Amsler grid testing is poor in measuring small central scotomata. In our series of patients, the absolute scotoma averaged 44.3 minutes of arc in radius, and we were able to detect a scotoma as small as 18.0 minutes of arc in one patient. We believe that our technique would be able to detect a scotoma as small as 7.0 minutes of arc in radius (the maximum radius of a 6 X 6-pixel square = 4.24 pixels, which in a 20 0 field would be equivalent to 7.0 minutes of arc). Using a 20 0 field, we have detected a scotoma as small as 16 minutes of arc in another patient with a small central macular scar (unpublished). We believe that microperimetry using the SLO provides the most sensitive means to detect and quantitate small scotomata in the macular area. This technique allows the examiner to assess and correct for fixation as well as make direct real-time correlation of visual function and anatomic pathology. In a previous series, Acosta et aC were able to demonstrate an absolute scotoma in the area of neurosensory defect in 26 of 26 eyes with macular holes. For this testing,
1515
Ophthalmology
Volume 100, Number 10, October 1993
Figure 2. A 6S-year-old man with an S -week history of reduced vision and metamorphopsia. Best-corrected vision was 20/40. A, red-free photograph demonstrates the neurosensory defect and surrounding neurosensory detachment. B, fluorescein angiogram demonstrates a window defect in the area of neurosensory defect. C, scanning laser ophthalmoscopic photograph shows the neurosensory defect and the surrounding area of darker discoloration. possibly delimiting the full extent of the neurosensory detachment. D, macular microperimetry shows the intensity isopters of scotomata of the macular hole and surrounding retina; the inner isopter is from 63 to 255. the next isopter is 31. and the outer isopter is 15.
the authors used a hybrid perimetry technique with the SLO which helped to control for fixation. Although a dense scotoma was found over the neurosensory defect in all eyes, only 2 of 26 eyes had a relative scotoma with detectable functional alterations at the margin of the hole or in the area of the surrounding neurosensory retinal detachment. Their technique used bleaching photopic conditions and flashing dark (relative to background illumination) test stimuli. Their smallest test stimulus was a 6 X6-pixel square, as in our study; however, the minimum contrast level used to check for a relative scotoma in their study was 40% (l 00 of 256 levels of "gray intensity"). Our technique differs, in that it is performed under mesopic conditions similar to kinetic testing using a Goldmann perimeter. Our test stimuli are bright on the relatively dim background. Brightness of the test stimuli is based on a scale of 256 gray levels. The software allows us to vary this from 0 to 255, and we elected to establish a log 2 scale of isopters, 2n - I. In our series, retinal
1516
dysfunction was demonstrated by the isopters of relative scotomata in the retina surrounding the neurosensory defect of the macular hole. Retina closer to the edge of the macular hole had more dysfunction and only perceived the brighter test stimuli. Dimmer test stimuli were first seen further from the hole, and normal retinal function was appreciated at the dimmest isopter intensity-l5. Often, the two dimmest isopters were outside the area of neurosensory defect, as perceived on clinical examination or in clinical photographs, demonstrating that retinal dysfunction is often more widespread than the apparent clinical pathology. It is unclear why this occurs; however, there may be more widespread neurosensory detachment than that detectable by clinical examination. Photographs with the SLO show a surrounding area of dark discoloration which is larger than the neurosensory detachment as detected on clinical examination, color photographs, or fluorescein angiography. In our series, the absolute scotoma exactly corresponded to the area of neurosen-
Sjaarda et al . Idiopathic Macular Holes
Figure 3. A comparison of two patients with good vision. A and B, scanning laser ophthalmoscopic photograph and macular microperimetry of a 57 -year-old woman with best-corrected visual acuity of 20/50-1. Notice larger absolute scotoma size (innermost circle) and smaller relative scotoma sizes. C and D, scanning laser ophthalmoscopic photograph and macular microperimetry of a 72-year-old woman with a best-corrected visual acuity of 20/50-2. Notice the small absolute scotoma size (innermost circle) and the larger surrounding relative scotoma sizes.
firmed in another series, 3 but was not found in a third. 6 In our study, there was an excellent correlation between visual acuity and the size of the macular hole and the size of the surrounding relative scotomata in the neurosensory
sory defect and therefore was a measure of macular hole size. Morgan and Schatz9 reported a close correlation of macular hole diameter and visual acuity. This was con-
20/25
iii a:
20/40
!!!
20/63
....
20/25
• • •• •
..•
0
~
~
•• ••
•
20/160 20/200
•
o
5
10
....
!!!
••
••••
•
::J
A
20/40
•
0
••
20/100
:5
iii a:
•
15
• •
~
~
• •
:5
20
20/100
::J
Radius of absolute scotoma (pixels)
20/63
25
B
•
20/160 20/200
35
30
••
•
• •• •••
45
55
65
•
-
••
• •
•
•
••
•
• •
•
75
85
95
•
105
115
Radius of 15 isopter scotoma (pixels)
Figure 4. Best-corrected visual acuity was correlated to the radius of the absolute scotoma and the 1S-intensity isopter scotoma by multivariate analysis, using the two as independent variables (P < 0.002). A, comparison of visual acuity and radius of absolute scotoma size. B, comparison of visual acuity and radius of 1S-intensity isopter relative scotoma.
1517
Ophthalmology
Volume 100, Number 10, October 1993
30
120~~~~~~~~~~~~~~~~~~~-,
I
•
01
......•................ --............._-................_...... .
• • •
.
• •• ::..
• •
.....
E
•
•
~0 ~
~
c.
•
0
J!}.
~
'0
]
a:
A
20
30
••
60
'" : •
40,"
•
•
• • • •
••
~
40
50
60
70
80
90
100
Symptom Duration (weeks)
20+···················································.........................................................................................................................................
··-1
O,+-~,-~,-~,-~,-~,-~,-~,-~r-~r-~
o+-~,-~,-~,-~,-~,-~.-~,-~,-~,-~
10
•
I
Ul ::l
5 .............................. --...........~.............-..-..........----.......... --..............................................................._......
o
_.
•
•
80·....··············..·······..···....· ..· .. ·... ·
CJ)
•
•
•
•
1ii' 100~········································.!c.:·····..........................................................................................................................................
........................................................... ___ .......................... __.............J!!!II.. ....
8
o
10
20
30
40
50
60
70
80
90
100
Symptom Duration (weeks)
Figure 5. The size of the absolute scotoma (A) and the 1S-intensity isopter relative scotoma (B) were correlated with the length of the patients' duration of symptoms (P < 0.05).
detachment. In general, patients with poor vision have larger holes and larger surrounding areas of neurosensory detachment with relative scotomata (Fig 1). Patients who retain better vision demonstrate that the visual loss is a function of both the neurosensory defect size and the neurosensory detachment size. In Figure 3, the comparison of two patients with 20/50 visual acuity is shown. The first patient shows a larger macular hole with a larger absolute scotoma, but smaller relative scotomata in the surrounding neurosensory detachment. The second patient has a smaller absolute scotoma and neurosensory defect, but larger relative scotomata in the surrounding neurosensory detachment. Further studies may better define what significance these findings might have toward predicting the outcome of macular hole surgery. In our series there also was a high correlation between duration of symptoms and the size of the macular hole and the surrounding relative scotomata. Eyes with longer duration of symptoms had larger holes and larger areas of surrounding retinal dysfunction. Our series did not provide for longitudinal follow-up of patients; however, other reports have shown that the neurosensory defect, as well as the neurosensory detachment, tends to enlarge with time. 5,9 Hole enlargement may have important implications for outcome after surgery for macular holes. In conclusion, using microperimetry with the SLO, we have demonstrated absolute and relative scotomata in 30 of 30 eyes with idiopathic macular holes in a consecutive series of30 patients. Best-corrected visual acuity correlated with the size of the neurosensory defect and neurosensory detachment, conclusively demonstrating that the visual loss associated with full-thickness macular holes is due to the reduction of retinal function in the area of the surrounding neurosensory detachment as well as to the ab-
1518
sence of retinal function in the area of neurosensory defect. Future studies of microperimetry of macular holes before and after surgery may help in our understanding of the mechanisms by which visual acuity improves after successful closure or flattening of macular holes by vitreous surgical techniques.
References 1. Kelly NE, Wendel RT. Vitreous surgery for idiopathic macular holes. Results of a pilot study. Arch Ophthalmol 1991; 109:654-9. 2. Glaser BM, Michels RG, Kuppermann BD, et al. Transforming growth factor-J32 for the treatment offull-thickness macular holes. A prospective randomized study. Ophthalmology 1992;99:1162-73. 3. Gass JDM. Idiopathic senile macular hole: its early stages and pathogenesis. Arch Ophthalmol 1988;106:629-39. 4. Aaberg TM. Macular holes: a review. Surv Ophthalmol 1970;15:139-62. 5. Johnson RN, Gass JDM. Idiopathic macular holes: observations, stages of formation, and implications for surgical intervention. Ophthalmology 1988;95:917-24. 6. Smith RG, Hardman Lea SJ, Galloway NR. Visual performance in idiopathic macular holes. Eye 1990;4: 190-4. 7. Acosta F, Lashkari K, Reynaud X, et al. Characterization of functional changes in macular holes and cysts. Ophthalmology 1991;98:1820-3. 8. Early Treatment Diabetic Retinopathy Study Research Group. Early Treatment Diabetic Retinopathy Study design and baseline patient characteristics. ETDRS report number 7. Ophthalmology 1991 ;98(Suppl):7 41-56. 9. Morgan CM, Schatz H. Involutional macular thinning. A pre-macular hole condition. Ophthalmology 1986;93: 15361.
With the advent of promising new vitreous surgical techniques to treat and restore vision in eyes with macular
Originally received: December 22, 1992. Revision accepted: March 8, 1992. From the Retina Institute of Maryland, Baltimore, Maryland. Presented in part at the Annual Meeting of the Vitreous Society, Laguna Niguel, October 1992. Reprint requests to Raymond N. Sjaarda, MD, Retina Institute of Maryland, O'Dea Medical Arts Bldg, 7505 Osler Dr, Suite 103, Baltimore, MD 21204.
holes 1,2 there has been renewed interest in better characterizing precisely the associated visual alterations. Visual loss associated with macular holes is thought to be related to the neurosensory defect as well as the surrounding neurosensory detachment. 3- s However, previous reports have had limited success in demonstrating absolute and/or relative scotomata corresponding to the macular holes,s-7 and have not shown any correlation of these to visual acuity. We present our findings of absolute and relative scotomata as demonstrated by microperimetry using the scanning laser ophthalmoscope (SLO) in a consecutive series of 30 patients with idiopathic macular holes.
1513
Ophthalmology
Volume 100, Number 10, October 1993
Patients and Methods A total of 30 eyes of 30 consecutive patients with idiopathic macular holes and symptoms of reduced vision were studied. Only eyes with 1+ nuclear sclerosis or less were included in this study group. Best-corrected visual acuity was obtained by protocol refraction and measurement by a trained clinical coordinator using the logarithmic eye chart used in the Early Treatment of Diabetic Retinopathy Study.8 Visual acuities were converted to visual acuity scores for statistical comparison (see below). 8 All patients underwent examination, including slitlamp biomicroscopic, funduscopic, and fundus biomicroscopic examination with contact lens or 78-diopter (D) indirect condensing lens. Fundus photography and fluorescein angiography were performed. The diagnosis of a macular hole was made by the presence of a full-thickness neurosensory defect and the presence of a window defect on fluorescein angiography. Holes were staged according to the classification proposed by Gass. 3.5 Microperimetry was performed on all patients using the Rodenstock SLO (Rodenstock, USA, Danbury, CT) and a software package developed for microperimetry using this instrument (Scanning Laser Ophthalmoscope Psychophysics Programs, Eye Research Institute of Retina Foundation, Boston, MA). All microperimetry was performed by two examiners (RNS and OAF). The SLO unit had helium-neon, argon blue-green, and infrared lasers. To provide for optimal imaging as well as mesopic background illumination, the helium-neon laser was adjusted to an output of 1 to 2 poW and the infrared laser was adjusted for an output of 20 poW. The argon laser was not used. For all tests, 40° fields were used. Fixation intensity as well as isopters of stimulus intensity were established in the following manner. Contrast of fixation and test stimuli can be adjusted by the software from 0% to 100% in 256 increments (0 through 255) based on an 8-bit computer microprocessor (2 8 = 256) producing "shades of gray." For fixation, a 100% contrast (255) cross with a size of 10 X 10 to 30 X 30 pixels was used. For a test stimulus, a 6 X6-pixel square (equivalent to 19.7 minutes of arc square, corresponding to approximately 97 porn square on the retina) was used for kinetic testing. A total of five contrast settings were used for the test stimulus to create five isopters for kinetic testing. The five contrast settings of 15,31,63,127, and 255 vary by log 2 contrast intensity (2n - 1), thus creating logarithmic increments of intensity. Kinetic perimetry was performed similar to that of testing with a Goldmann kinetic perimeter. The examiner is able to view the macular pathology and assess fixation in "real time" while performing the micro perimetry. Patients were asked to fixate on the fixation cross and signal when the test stimulus was first seen. lsopters were mapped on an acetate overlay on the video monitor and subsequently mapped out using the perimetry software. The software provides digital analysis which converts the area mapped into pixels. Statistical comparisons were performed using multivariate analysis (for vision and scotoma comparisons) or linear regression (for scotoma and duration of symptom comparisons).
1514
Results
------------------------------There were 20 women and 10 men in the study. The
average age was 64.6 years (range, 53-75 years). The average best-corrected vision was 20/80-3 (range, 20/4020/200+2). All patients noted symptoms ofloss of vision and/or metamorphopsia. The average duration of symptoms before presentation was 33.7 weeks (range, 1-100 weeks). There were 5 stage 2 holes, 18 stage 3 holes, and 7 stage 4 holes. Absolute and relative scotomata were demonstrated in all patients. All patients showed an absolute scotoma corresponding exactly to the size of the neurosensory defect. Retina at the rim of the macular hole perceived the brightest (255) intensity stimulus in all eyes. The 127intensity stimulus was perceived in 12 of 30 eyes by the retina at the margin of the hole. In the remaining 18 patients, the 127-intensity isopter was seen further away from the margin of the hole, creating a surrounding isopter of relative scotoma. All patients showed additional surrounding scotomata of progressively dimmer isopters (Figs 1 and 2). These scotomata were generally within the area of visible neurosensory detachment. However, the two dimmest isopters, the 31- and 15-intensity levels, often were located outside the area of apparent neurosensory detachment. Vision was a function of both the absolute scotoma and the surrounding relative scotoma (Fig 3). Best-corrected visual acuity was correlated to the size of the absolute and the 15-intensity isopter relative scotoma by multivariate analysis using the two as independent variables (P < 0.002). The average radius of the absolute scotoma was 13.5 pixels (44.3 minutes of arc, or 219 porn on the retina). The minimum and maximum radii were 5.5 pixels (18.0 minutes of are, or 89 porn on the retina) and 25.4 pixels (83.3 minutes of are, or 412 porn on the retina), respectively (Fig 4). The average, minimum, and maximum radii of the 15-intensity relative scotoma were 70.4 pixels (230.9 minutes of are, or 1142 porn on the retina), 37.3 pixels (122.3 minutes of are, or 605 porn on the retina), and 112.2 pixels (368.0 minutes of arc, or 1820 porn on the retina), respectively (Fig 4). Figure 5A shows a comparison of the size of absolute scotoma (which equals the size of the neurosensory defect) and duration of symptoms. Eyes with a longer duration of symptoms, and presumably older holes, had significantly larger holes (R 2 = 0.40; P < 0.05). Figure 5B shows a comparison of the size of the 15-intensity isopter relative scotoma with the duration of the patients' symptoms. Eyes with a longer duration of symptoms showed significantly larger radii of the 15 isopter scotoma (R2 = 0.37; P < 0.05).
Discussion Using kinetic microperimetry with the SLO, we demonstrated absolute and relative scotomata in 30 eyes of 30 consecutive patients with idiopathic macular holes. In addition, we validated the logical hypothesis that visual loss in this condition is related to the absence of retinal func-
Sjaarda et al . Idiopathic Macular Holes
Figure 1. A 63-year-old woman with a 44-week history of poor vision in her left eye is shown. Best-corrected visual acuity was 20/80-2. A, red-free photograph shows a full-thickness stage 3 macular hole and the surrounding rim of a neurosensory detachment. B, fluorescein angiogram demonstrates a window defect in the area of a neurosensory detachment. C, scanning laser ophthalmoscopic photograph shows neurosensory defect and surrounding area of darker discoloration, possibly delimiting the full extent of the neurosensory detachment. D, macular microperimetry shows the intensity isopters of scotomata of the macular hole and surrounding retina; from inner to outer, they are 255 (absolute), 127, 63, 31, and 15.
tion in the area of neurosensory defect as well as the reduction of retinal function in the area of the surrounding neurosensory detachment. Previously described techniques using an Amsler grid 5,6 or hybrid forms of microperimetry4 have demonstrated absolute and/or relative scotomata with limited consistency. Johnson and Gass5 demonstrated a scotoma, or a scotoma with metamorphopsia, on Amsler grid testing in as many as 40% of 109 patients with stage 3 or 4 macular holes. Smith et al,6 using Amsler grid techniques, demonstrated an absolute scotoma in 8 of 24 eyes of patients with full-thickness macular holes. They calculated the scotoma to be less than 30 in diameter in these patients. No absolute scotoma could be perceived in 16 of24 eyes, but all of these eyes demonstrated metamorphopsia. They concluded that the predicted central scotoma would be less than lOin diameter « 1 square on an Amsler grid) and therefore would not be recognized. They also concluded that accurate plotting of small central scotomata is problematic because of the difficulty in maintaining
fixation. Weare in agreement that Amsler grid testing is poor in measuring small central scotomata. In our series of patients, the absolute scotoma averaged 44.3 minutes of arc in radius, and we were able to detect a scotoma as small as 18.0 minutes of arc in one patient. We believe that our technique would be able to detect a scotoma as small as 7.0 minutes of arc in radius (the maximum radius of a 6 X 6-pixel square = 4.24 pixels, which in a 20 0 field would be equivalent to 7.0 minutes of arc). Using a 20 0 field, we have detected a scotoma as small as 16 minutes of arc in another patient with a small central macular scar (unpublished). We believe that microperimetry using the SLO provides the most sensitive means to detect and quantitate small scotomata in the macular area. This technique allows the examiner to assess and correct for fixation as well as make direct real-time correlation of visual function and anatomic pathology. In a previous series, Acosta et aC were able to demonstrate an absolute scotoma in the area of neurosensory defect in 26 of 26 eyes with macular holes. For this testing,
1515
Ophthalmology
Volume 100, Number 10, October 1993
Figure 2. A 6S-year-old man with an S -week history of reduced vision and metamorphopsia. Best-corrected vision was 20/40. A, red-free photograph demonstrates the neurosensory defect and surrounding neurosensory detachment. B, fluorescein angiogram demonstrates a window defect in the area of neurosensory defect. C, scanning laser ophthalmoscopic photograph shows the neurosensory defect and the surrounding area of darker discoloration. possibly delimiting the full extent of the neurosensory detachment. D, macular microperimetry shows the intensity isopters of scotomata of the macular hole and surrounding retina; the inner isopter is from 63 to 255. the next isopter is 31. and the outer isopter is 15.
the authors used a hybrid perimetry technique with the SLO which helped to control for fixation. Although a dense scotoma was found over the neurosensory defect in all eyes, only 2 of 26 eyes had a relative scotoma with detectable functional alterations at the margin of the hole or in the area of the surrounding neurosensory retinal detachment. Their technique used bleaching photopic conditions and flashing dark (relative to background illumination) test stimuli. Their smallest test stimulus was a 6 X6-pixel square, as in our study; however, the minimum contrast level used to check for a relative scotoma in their study was 40% (l 00 of 256 levels of "gray intensity"). Our technique differs, in that it is performed under mesopic conditions similar to kinetic testing using a Goldmann perimeter. Our test stimuli are bright on the relatively dim background. Brightness of the test stimuli is based on a scale of 256 gray levels. The software allows us to vary this from 0 to 255, and we elected to establish a log 2 scale of isopters, 2n - I. In our series, retinal
1516
dysfunction was demonstrated by the isopters of relative scotomata in the retina surrounding the neurosensory defect of the macular hole. Retina closer to the edge of the macular hole had more dysfunction and only perceived the brighter test stimuli. Dimmer test stimuli were first seen further from the hole, and normal retinal function was appreciated at the dimmest isopter intensity-l5. Often, the two dimmest isopters were outside the area of neurosensory defect, as perceived on clinical examination or in clinical photographs, demonstrating that retinal dysfunction is often more widespread than the apparent clinical pathology. It is unclear why this occurs; however, there may be more widespread neurosensory detachment than that detectable by clinical examination. Photographs with the SLO show a surrounding area of dark discoloration which is larger than the neurosensory detachment as detected on clinical examination, color photographs, or fluorescein angiography. In our series, the absolute scotoma exactly corresponded to the area of neurosen-
Sjaarda et al . Idiopathic Macular Holes
Figure 3. A comparison of two patients with good vision. A and B, scanning laser ophthalmoscopic photograph and macular microperimetry of a 57 -year-old woman with best-corrected visual acuity of 20/50-1. Notice larger absolute scotoma size (innermost circle) and smaller relative scotoma sizes. C and D, scanning laser ophthalmoscopic photograph and macular microperimetry of a 72-year-old woman with a best-corrected visual acuity of 20/50-2. Notice the small absolute scotoma size (innermost circle) and the larger surrounding relative scotoma sizes.
firmed in another series, 3 but was not found in a third. 6 In our study, there was an excellent correlation between visual acuity and the size of the macular hole and the size of the surrounding relative scotomata in the neurosensory
sory defect and therefore was a measure of macular hole size. Morgan and Schatz9 reported a close correlation of macular hole diameter and visual acuity. This was con-
20/25
iii a:
20/40
!!!
20/63
....
20/25
• • •• •
..•
0
~
~
•• ••
•
20/160 20/200
•
o
5
10
....
!!!
••
••••
•
::J
A
20/40
•
0
••
20/100
:5
iii a:
•
15
• •
~
~
• •
:5
20
20/100
::J
Radius of absolute scotoma (pixels)
20/63
25
B
•
20/160 20/200
35
30
••
•
• •• •••
45
55
65
•
-
••
• •
•
•
••
•
• •
•
75
85
95
•
105
115
Radius of 15 isopter scotoma (pixels)
Figure 4. Best-corrected visual acuity was correlated to the radius of the absolute scotoma and the 1S-intensity isopter scotoma by multivariate analysis, using the two as independent variables (P < 0.002). A, comparison of visual acuity and radius of absolute scotoma size. B, comparison of visual acuity and radius of 1S-intensity isopter relative scotoma.
1517
Ophthalmology
Volume 100, Number 10, October 1993
30
120~~~~~~~~~~~~~~~~~~~-,
I
•
01
......•................ --............._-................_...... .
• • •
.
• •• ::..
• •
.....
E
•
•
~0 ~
~
c.
•
0
J!}.
~
'0
]
a:
A
20
30
••
60
'" : •
40,"
•
•
• • • •
••
~
40
50
60
70
80
90
100
Symptom Duration (weeks)
20+···················································.........................................................................................................................................
··-1
O,+-~,-~,-~,-~,-~,-~,-~,-~r-~r-~
o+-~,-~,-~,-~,-~,-~.-~,-~,-~,-~
10
•
I
Ul ::l
5 .............................. --...........~.............-..-..........----.......... --..............................................................._......
o
_.
•
•
80·....··············..·······..···....· ..· .. ·... ·
CJ)
•
•
•
•
1ii' 100~········································.!c.:·····..........................................................................................................................................
........................................................... ___ .......................... __.............J!!!II.. ....
8
o
10
20
30
40
50
60
70
80
90
100
Symptom Duration (weeks)
Figure 5. The size of the absolute scotoma (A) and the 1S-intensity isopter relative scotoma (B) were correlated with the length of the patients' duration of symptoms (P < 0.05).
detachment. In general, patients with poor vision have larger holes and larger surrounding areas of neurosensory detachment with relative scotomata (Fig 1). Patients who retain better vision demonstrate that the visual loss is a function of both the neurosensory defect size and the neurosensory detachment size. In Figure 3, the comparison of two patients with 20/50 visual acuity is shown. The first patient shows a larger macular hole with a larger absolute scotoma, but smaller relative scotomata in the surrounding neurosensory detachment. The second patient has a smaller absolute scotoma and neurosensory defect, but larger relative scotomata in the surrounding neurosensory detachment. Further studies may better define what significance these findings might have toward predicting the outcome of macular hole surgery. In our series there also was a high correlation between duration of symptoms and the size of the macular hole and the surrounding relative scotomata. Eyes with longer duration of symptoms had larger holes and larger areas of surrounding retinal dysfunction. Our series did not provide for longitudinal follow-up of patients; however, other reports have shown that the neurosensory defect, as well as the neurosensory detachment, tends to enlarge with time. 5,9 Hole enlargement may have important implications for outcome after surgery for macular holes. In conclusion, using microperimetry with the SLO, we have demonstrated absolute and relative scotomata in 30 of 30 eyes with idiopathic macular holes in a consecutive series of30 patients. Best-corrected visual acuity correlated with the size of the neurosensory defect and neurosensory detachment, conclusively demonstrating that the visual loss associated with full-thickness macular holes is due to the reduction of retinal function in the area of the surrounding neurosensory detachment as well as to the ab-
1518
sence of retinal function in the area of neurosensory defect. Future studies of microperimetry of macular holes before and after surgery may help in our understanding of the mechanisms by which visual acuity improves after successful closure or flattening of macular holes by vitreous surgical techniques.
References 1. Kelly NE, Wendel RT. Vitreous surgery for idiopathic macular holes. Results of a pilot study. Arch Ophthalmol 1991; 109:654-9. 2. Glaser BM, Michels RG, Kuppermann BD, et al. Transforming growth factor-J32 for the treatment offull-thickness macular holes. A prospective randomized study. Ophthalmology 1992;99:1162-73. 3. Gass JDM. Idiopathic senile macular hole: its early stages and pathogenesis. Arch Ophthalmol 1988;106:629-39. 4. Aaberg TM. Macular holes: a review. Surv Ophthalmol 1970;15:139-62. 5. Johnson RN, Gass JDM. Idiopathic macular holes: observations, stages of formation, and implications for surgical intervention. Ophthalmology 1988;95:917-24. 6. Smith RG, Hardman Lea SJ, Galloway NR. Visual performance in idiopathic macular holes. Eye 1990;4: 190-4. 7. Acosta F, Lashkari K, Reynaud X, et al. Characterization of functional changes in macular holes and cysts. Ophthalmology 1991;98:1820-3. 8. Early Treatment Diabetic Retinopathy Study Research Group. Early Treatment Diabetic Retinopathy Study design and baseline patient characteristics. ETDRS report number 7. Ophthalmology 1991 ;98(Suppl):7 41-56. 9. Morgan CM, Schatz H. Involutional macular thinning. A pre-macular hole condition. Ophthalmology 1986;93: 15361.