- Email: [email protected]
Involutional Diabetic Retinopathy
INVOLUTIONAL DIABETIC RETINOPATHY WILLIAM J. RAMSAY, M.D., ROBERT C. RAMSAY, M.D., RICHARD L. P U R P L E , P H . D . , AND WILLIAM H. KNOBLOCH, M.D.
Minneapolis, Minnesota 1,2
Studies show spontaneous regression of ocular neovascularization has occurred late in the course of proliferative diabetic retinopathy. The resultant fundus appear ance is characterized by vascular attenua tion, optic nerve pallor, pigmentary dis persion, and replacement of neovasculari zation by avascular glial tissue (Fig. 1). This is referred to as the atrophie, burned-out, or involutional phase of dia betic retinopathy. One of us (W.H.K.) observed that patients with this end-stage proliferative retinopathy complain of dif ficulties with night vision and, addition ally, that there is a superficial resem blance to pigmentary degeneration of the retina. These observations suggest that despite the clinical improvement in the late stage, as compared to the earlier ac tive phases of proliferative retinopathy, abnormalities of retinal function in such eyes should be demonstrable by electrophysiologic testing methods. Additionally, the frequent occurence of involutional retinopathy after vitrectomy surgery3 makes further study of this phase increasingly important. SUBJECTS AND METHODS
We studied 12 patients with the clinical appearance of involutional diabetic reti nopathy. Patients who had previously reFrom the Departments of Ophthalmology (Drs. W. J. Ramsay, R. C. Ramsay, and Knobloch) and Physiology (Dr. Purple), University of Minnesota, Minneapolis, Minnesota. This study was supported in part by research grants EY00293 and EY02316-01 from the National Eye Institute and the Minnesota Lions Club. This study was presented in part before the Association for Research in Vision and Ophthal mology Annual Meeting, Sarasota, Florida, April 26, 1977. Reprint requests to William H. Knobloch, M.D., Box 493 Mayo Bldg., Department of Ophthalmology, University of Minnesota, Minneapolis, ΜΝ 55455.
Fig. 1 (Ramsay and associates). Posterior pole photograph of involutional retinopathy showing optic nerve pallor, residual avascular gliosis, arteriolar and venous attenuation, pigmentary mottling, and contracture of the internal limiting membrane of the retina.
ceived extensive photocoagulation were excluded from this study. All patients had previously documented neovasculariza tion or vitreous hemorrhage and, at the time of this study, had clear media and no evidence of active proliferative disease. The patients ranged in age from 30 to 66 years (mean, 44 years), and all had insulin-dependent diabetes mellitus. The duration of diabetes ranged from 18 to 40 years (mean, 27 years). Five eyes had severe visual loss (hand movements or no light perception) secondary to traction retinal detachments and were not studied. The following determinations were made on the remaining 19 eyes: 1. Best corrected visual acuity using the methods developed by the Diabetic Retinopathy Study protocol.4 2. Goldmann visual field.
851
852
AMERICAN JOURNAL OF OPHTHALMOLOGY
3. Fundus photography of the seven standard fields identified in the Diabet ic Retinopathy Study protocol. 4 4. Dark adaptation curve using the Goldmann-Weekers dark adaptometer. 5. Fluorescein angiography. 6. Color vision using the Farnsworth-Dichotomous test (Panel D-15), the Farnsworth-Munsell 100-Hue test, and the Nagel anomaloscope. 7. Electroretinography (ERG) was performed after pupil dilatation, and initial testing was done after light adap tation at greater than 5 foot-lamberts. A Burian-Allen bipolar contact lens was used with a Grass PS-22 photostimulator placed 12 inches above the patient's eye. The stimulus measured approxi mately 30 degrees of visual field. Photopic recordings were made by using a series of photopically balanced white, red, and blue lights. Providing there was no history of a seizure disorder, the 30-Hz photopic flicker test was done to determine the phase responses to white light at maximum intensity (16 on PS 22, with no filters). In our previous control studies, we used the PS-22 photostimulator with 30 degrees of reti na illuminated and confirmed Berson and associates' 5 - 7 results regarding phase relationships obtained with a ganzfeld stimulus in the presence of a progressive tapetoretinal degeneration. Our patients were then dark-adapted for 20 minutes and the scotopic re sponses to white light flashes were re corded. Starting with the threshold in tensity necessary to evoke a measurable scotopic b-wave (10 to 20 μλθ, the in tensity was increased in steps of 0.5 log units or less in order to define the maxi mal b-wave. As b-wave amplitude was measured from the ERG baseline to the peak, care was taken in the critical area (usually between 2.0 and 2.5 log units of light attenuation) by proceeding in
DECEMBER, 1977
0.1 or 0.2 log unit increments to deter mine that point at which the scotopic a-wave began to diminish the peak b wave amplitude. After completing the scotopic series of recordings, we used deep red room light to allow transfer of the contact lens to the fellow eye. After an addition al three minutes of dark adaptation, the scotopic series was repeated. If neces sary, the patient was then light-adapted for ten minutes and the photopic series was repeated. Each eye was tested sep arately as our patients tolerated separate testing better than testing of both eyes simultaneously. Using Polaroid photographs of the oscilloscope tracings, we made meas urements from waves recorded. A-wave amplitude was measured from the ERG baseline to the trough. B-wave ampli tude was measured from the ERG base line to the peak. The latency to response onset (measured on 5 or 10 msec per division of sweep speed) and implicit time to the peak of the response were also determined. 8. Electro-oculography (EOG) was performed by recording each minute the potentials produced by ocular rota tions of 60 degrees for 15 minutes dur ing dark adaptation and then for 15 minutes during light adaptation. The light-peak (L) to dark-trough (D) ratio was then determined. The lower limit of normal for the L/D ratio in our labo ratory is 1.60. The EOG curve for each patient was plotted as microvolts per degree of ocular rotation vs time. RESULTS
The best corrected visual acuity ob tained was 6/6 (20/20) in eight eyes, 6/9 (20/30) in five eyes, 6/12 (20/40) in three eyes, 6/15 (20/50) in one eye, 6/30 (20/ 100) in one eye, and 6/60(20/200) in one eye. One eye was aphakic and all other
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
eyes had minimal refractive errors withthe spherical equivalents ranging from +0.50 to - 3 . 0 0 diopters. The visual field abnormalities recorded can be divided into two groups. One group of seven eyes showed mild abnor malities consisting of mild to moderate constriction of all isopters tested. The remaining 12 eyes had more severe abnor malities and fell into the second group. Each of these eyes had large, dense, nerve-fiber defects. Associated defects found included isolated scotomata in four eyes and a partial ring scotoma in an additional four eyes. The single eye with visual acuity of 6/60 (20/200) had the most severe visual field abnormality con sisting of a dense ring scotoma with a remaining central field of less than 10 degrees to the IV/4 target. A dark adaptation curve was deter mined in 14 eyes. Four eyes showed a delayed rod-cone break, as well as an abnormal final rod threshold which re mained elevated above a log intensity of 10 4 . The remaining ten eyes showed a delay in the rod-cone break but reached normal final rod threshold levels. The macular angiographie findings were similar in all eyes studied (Fig. 2). The abnormalities included delayed ar terial flow, areas of capillary obliteration, and minimal intraretinal vascular anoma lies. Additionally, thinning and irregular ity of the retinal pigment epithelium pro duced a mottled fluorescein pattern with window defects. No eye showed active proliferative disease, although avascular glial tissue was present along the superior and inferior vascular arcades in several eyes. The E O G was recorded in 16 eyes (Fig. 3) and all showed subnormal recordings with the L/D ratio ranging from 1.00 to 1.56 (mean, 1.33). Electroretinography was done in 17 of the 19 eyes included in this study (Fig. 4).
853
Fig. 2 (Ramsay and associates). Fluorescein angiography demonstrates delayed and incomplete arteriolar flow, capillary obliteration, and retinal pig ment epithelial atrophy with window defects.
The recordings indicated reduced a-wave and b-wave amplitudes under both photopic and scotopic testing conditions. The maximum a-wave obtained ranged from 70 μ ν to 300 μ ν with a mean of 167 μ ν . T h e maximum scotopic b-wave obtained ranged from 60 μΥ to 300 μΥ with a mean
DARK ! LIGHT
1.41
0 4
12 16 20 24 28
J I J _ J LU_ 0 4 8 12 16 20 24 28
Time in Minutes Fig. 3 (Ramsay and associates). Left, Normal curves. Right, Representative depressed curves in eyes with involutional retinopathy.
854
AMERICAN JOURNAL OF OPHTHALMOLOGY NORMAL
PROLIFERA-
IVE
INVOLUT ZONAL
PHOTOPIC 30 HZ FUCKBR
30 HZ FUCKER
SCOTOPIC
^__
1 100 miemirclts
*~* IO ora20 millistcortdi
Fig. 4 (Ramsay and associates). Photopic respons es to red, blue, and white light flashes obtained in one normal control eye and two eyes with involutional retinopathy. The response to the 30-Hz flicker (stimulus flash indicated by arrows) is abnormal in the involutional eyes. Scotopic responses to variable intensity white light flashes show marked reduction in the maximum b-wave amplitude in the involu tional eyes.
DECEMBER, 1977
of 130 μν. Normal values in our laborato ry are 200 to 500 μν for both the peak scotopic a-wave and the peak scotopic b-wave. Additionally, the latency of response to white light under scotopic conditions was delayed in all 17 eyes tested. As well, the photopic 30-Hz flicker test showed an abnormal phase relation and subnormal amplitudes in all eyes tested. Finally, b-wave oscillations were observable, but depressed in 15 eyes and absent to the highest light intensity flash in two eyes. Abnormalities of color vision were found in eight of the nine patients tested. Six patients showed evidence of an ac quired dyschromatopsia (Fig. 5) with major involvement of the tritan axis. The Famsworth-Munsell 100 Hue test was the most diagnostic. All six patients showed pathologic total error scores ranging from 145 to 471 and located around the blueyellow axis.
Fig. 5 (Ramsay and associates). In four patients. Results of Fams worth-Munsell 100 Hue tests indi cated an acquired blue-yellow dys chromatopsia.
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
855
TEST-
' ,ϊ RETEST-
I
i 12 M
Fig. 6 (Ramsay and associates). Results of Farnsworth-Munsell 100 Hue test (two trials averaged) and D-15 Panel test in one patient with involutional retinopathy are consistent with diagnosis of acquired tritanopia.
The remaining two patients showed abnormalities consistent with a diagnosis of tritanopia or tritanomaly (Fig. 6). In both cases the Raleigh Equation was within the normal range. The D-15 Panel test was normal in one patient and showed dichromatic tritanopia in the second pa tient. The Farnsworth-Munsell 100 Hue test was abnormal for both and showed a marked tritan axis with few errors scat tered outside the tritan axis. DISCUSSION
We have developed our understanding of the natural course of proliferative dia betic retinopathy through several clinical studies. 1, 8~12 Once diabetic retinal neovascularization is present, a cycle devel ops characterized by the initial growth of
the new vessels, followed by regression. Dobree 1 refers to this cycle as the primary changes in proliferative diabetic retinop athy. The final visual outcome is modi fied by secondary changes occurring in the disk, retina, vitreous, choroid, and the anterior segment of the eye, resulting from hemorrhage or connective tissue for mation in the intraocular tissues.2 The present concept of the pathogenesis of diabetic retinopathy, both preproliferative and proliferative, is based on gradually increasing retinal ischemia. 13 ' 14 While the initiating factor or factors re main unclear, capillary obliteration, shunt vessel formation, and microaneurysm formation adjacent to avascular areas represent the early histopathologic findings.16"21 Increasing retinal ischemia
856
AMERICAN JOURNAL OF OPHTHALMOLOGY
at some critical point is thought to result in neovascularization, perhaps through the elaboration of a vasoproliferative fac tor. The late regression of neovasculariza tion probably results from further vascu lar insufficiency, leading to ischemia too severe to stimulate or support vasoproliferation. Spontaneous regression of neovascular ization in the diabetic eye results in im provement of visual prognosis. The eye is generally protected from catastrophic vit reous hemorrhage and visual acuity stabi lizes, frequently at a socially useful level (vision ä 6/15 [20/50]), Despite these beneficial effects, this study clearly shows that severe functional disturbances exist in such eyes. The marked reduction in the maximum b-wave amplitude of the ERG and the depression of the EOG indicate widespread cellular abnormalities affect ing the outer layers of the retina and the retinal pigment epithelium. Additionally, all eyes studied had de layed implicit times and an abnormal phase relation to the 30-Hz photopic flicker test. As these abnormalities are characteristic of tapetoretinal degenera tions, 6 ' 7 they suggest that progressive de terioration of retinal function occurs in eyes with involutional retinopathy. While the rate of deterioration has not yet been determined, the similarity to the ERG in dominantly inherited retinitis pigmentosa suggests that several decades could elapse before the ERG becomes nonrecordable and the eye blind. We have observed several patients with a history of diabetes mellitus of 35 to 45 years duration, ad vanced involutional retinopathy, and optic nerve atrophy, with reduction of visual acuity to the level of light percep tion. Furthermore, the variation in both the depression of the ERG b-wave ampli tude and the visual field abnormalities in the eyes studied support the concept that the involutional phase of proliferative di abetic retinopathy is not an end-stage, but represents a continuum of slow deteriora
DECEMBER, 1977
tion of retinal function that may culmi nate in blindness. In this study, the extent of cellular loss, measured by the reduction of the ERG b-wave amplitude, is similar to that achieved therapeutically following panretinal photocoagulation. 22 Additional ly, when extensive photocoagulation is effective in causing regression of neo vascularization the resultant clinical ap pearance is similar to the spontaneously occurring involutional phase of diabetic retinopathy with vascular attenuation and optic nerve pallor. The ability of panretinal photocoagulation to reduce signif icantly the incidence of severe visual loss in eyes with advanced proliferative dia betic retinopathy has been demonstrated by the Cooperative Diabetic Retinopathy Study. 23 However, while Frank 22 has shown no further deterioration of the ERG response up to six weeks after exten sive photocoagulation as compared to the response immediately after treatment, the long-term effects have not been studied. If further deterioration of retinal function will occur with time, as this study sug gests, it may be necessary to evaluate the long-term effect of extensive photocoagu lation treatment. We did not expect the characteristic and repeatable tritan color vision abnormality found in eight of nine patients studied. We suppose this acquired blue-yellow dyschromatopsia is associated with the optic nerve atrophy present in all cases, rather than with a specific photoreceptor abnormality, although the findings do not distinguish between these two possibili ties. The delimited tritan defects are more characteristic of a congenital color defi ciency than of acquired defects, which generally show widespread error scores in the red-green as well as the blue-yellow axes. 24 · 25 While most patients in this study noted slow dark adaptation clinically, we could document this in only four of the 14 eyes tested. Thus, this test of rod function
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
seems affected later in the course of invol-
utional retinopathy than in that of the tapetoretinal degenerations. These results indicate that, despite ex cellent central visual acuity (17 of 19 eyes â 6/15 (20/50) and improved clinical ap pearance, eyes with involutional retinop athy show marked functional abnormali ties. We have demonstrated involvement of the retina, retinal pigment epithelium, and optic nerve; the atrophie appearance of the fundus suggests involvement of the choroidal vasculature as well. The results support the concept that involutional reti nopathy represents a continuation of the diabetic ischemie process, rather than a phase of spontaneous improvement from the earlier active proliferative phase of the disease. The data presented here suggest that the involutional phase of proliferative di abetic retinopathy may result in a pro gressive loss of retinal function. Evi dence of progression has not yet been documented. To test this hypothesis, a lon gitudinal study has been devised at the University of Minnesota. Information con cerning progressive loss of retinal func tion, as well as the rate of deterioration, is important as renal transplantation, 26 hemodialysis, and other therapeutic modali ties are used to improve the long-term survival of patients developing the microvascular complications of diabetes melli tus. SUMMARY
The end-stage or involutional phase of proliferative diabetic retinopathy may re sult in stabilization of vision for long periods of time. However, the clinical resemblance to the progressive tapetore tinal degenerations suggests that marked functional impairment of the retina is present in such eyes. We studied 19 eyes with involutional retinopathy to document the status of the retinal function. Studies included fluo
857
rescein angiography, visual field exami nation, dark adaptation testing, color vi sion testing, electro-oculography and electroretinography (ERG). The results indicated marked function al abnormalities in all eyes. The ERG tracings showed uniformly subnormal re sponses and delayed implicit times, simi lar to those of dominantly inherited reti nal pigment degeneration, and indicative of a progressive retinal disorder. In two patients, color vision testing showed de fects similar to those seen in inherited tritanopia; and in the remaining patients, defects were indicative of an acquired blue-yellow dyschromatopsia. REFERENCES 1. Dobree, J. H.: Proliferative diabetic retinopa thy. Evolution of the retinal lesions. Br. J. Ophthalmol. 48:637, 1964. 2. Davis, M. D.: Vitreous contraction in prolifera tive diabetic retinopathy. Arch. Ophthalmol. 74:741, 1965. 3. Peyman, G. A., Huamonte, F., Locketz, A., and Goldberg, M. F.: Fluorescein angiography of the fundus after pars plana vitrectomy. Ann. Oph thalmol. 8:791, 1976. 4. Diabetic Retinopathy Study: Manual of Opera tions. Baltimore, Diabetic Retinopathy Study Coor dinating Center, 1972. 5. Berson, E. L., Gouras, P., and Hoff, M.: Tem poral aspects of the electroretinogram. Arch. Oph thalmol. 81:207, 1969. 6. Berson, E. L., Gouras, P., and Gunkel, R. D.: Rod responses in retinitis pigmentosa, dominantly inherited. Arch. Ophthalmol. 80:58, 1968. 7. : Progressive cone-rod degeneration. Arch. Ophthalmol. 80:68, 1968. 8. Burditt, A. F., Caird, F. I., and Draper, G. J.: The natural history of diabetic retinopathy. Q. J. Med. 37:303, 1968. 9. Davis, M. D.: The natural course of diabetic retinopathy. Trans. Am. Acad. Ophthalmol. Otolaryngol. 72:237, 1968. 10. Beethan, W. P.: Visual prognosis of proliferat ing diabetic retinopathy. Br. J. Ophthalmol. 47:611, 1963. 11. Davis, M.: Natural course of diabetic retinop athy. In Kimura, S. J., and Caygill, W. M. (eds.): Vascular Complications of Diabetes Mellitus with Special Emphasis on Microangiography of the Eye. St. Louis, C. V. Mosby, 1967, p. 139-169. 12. Dobree, J. H.: Evolution of lesions in prolife rative diabetic retinopathy. An 8-year photographic survey. In Goldberg, M., and Fine, S. (eds.): The Treatment of Diabetic Retinopathy. Washington, U.S. Public Health Service, 1969, p. 55-64.
858
AMERICAN JOURNAL OF OPHTHALMOLOGY
13. Davis, M., Myers, F., Engerman, R., DeVenecia, G., and Magli, Y.: Clinical observations con cerning the pathogenesis of dgabetic retinopathy. In Goldberg, M., and Fine, S. (eds.): The Treatment of Diabetic Retinopathy. Washington, U.S. Public Health Service, 1969, p. 47-53. 14. Goldberg, M. F.: The role of ischemia in the production of vascular retinopathies. In Lynn, J., Snyder, W., and Vaiser, A. (eds.): Diabetic Retinopa thy. New York,Gruneand Stratton, 1974, p. 47-63. 15. Cogan, D. G., and Kuwabara, T.; Capillary shunts in the pathogenesis of diabetes retinopathy. Diabetes 12:293, 1963. 16. Ashton, N.: Arteriolar involvement in diabetic retionpathy. Br. J. Ophthalmol. 37:282, 1953. 17. : Vascular basement membrane chang es in diabetic retinopathy. Br. J. Ophthalmol. 58: 344, 1974. 18. Ditzel, J.: Haemorheological factors in the development of diabetic microangiopathy. Br. J. Ophthalmol. 51:793, 1967. 19. Cogan, D. G., Troussaint, D., and Kuwabara, T.: Retinal vascular patterns. Arch. Ophthalmol. 66:100, 1961.
DECEMBER, 1977
20. Kuwabara, T., and Cogan, D. G.: Retinal vascular patterns. 6. Mural cells of the retinal capil laries: Arch. Ophthalmol. 69:114, 1963. 21. Yanoff, M.: Ocular pathology of diabetes mellitus. Am. J. Ophthalmol. 67:21, 1969. 22. Frank, N.: Visual fields and electroretinography following extensive photocoagulation. Arch. Ophthalmol. 93:591,1975. 23. Diabetic Retinopathy Study Research Group: Preliminary report on effects of photocoagulation therapy. Am. J. Ophthalmol. 81:383, 1976. 24. Krill, A. E., Smith, V. C , and Pokorny, J.: Further studies supporting the identity of congeni tal tritanopia and hereditary dominant optic atro phy. Invest. Ophthalmol. 10:182,1971. 25. —Similarities between congenital tritan de fects and dominate optic nerve atrophy. Coinci dence or identity. J. Optom. Soc. Am. 60:1132, 1970. 26. Najarian, J. S., Sutherland, D. E., Simmons, R. L., Howard, R. J., Kjellstrand, C. M., Mauer, S. M., Kennedy, W., Ramsay, R. C , Barbosa, J., and Goetz, F. C : Kidney transplantation for the uremie diabet ic. Surg. Gynecol. Obstet. 144:682, 1977.
Minneapolis, Minnesota 1,2
Studies show spontaneous regression of ocular neovascularization has occurred late in the course of proliferative diabetic retinopathy. The resultant fundus appear ance is characterized by vascular attenua tion, optic nerve pallor, pigmentary dis persion, and replacement of neovasculari zation by avascular glial tissue (Fig. 1). This is referred to as the atrophie, burned-out, or involutional phase of dia betic retinopathy. One of us (W.H.K.) observed that patients with this end-stage proliferative retinopathy complain of dif ficulties with night vision and, addition ally, that there is a superficial resem blance to pigmentary degeneration of the retina. These observations suggest that despite the clinical improvement in the late stage, as compared to the earlier ac tive phases of proliferative retinopathy, abnormalities of retinal function in such eyes should be demonstrable by electrophysiologic testing methods. Additionally, the frequent occurence of involutional retinopathy after vitrectomy surgery3 makes further study of this phase increasingly important. SUBJECTS AND METHODS
We studied 12 patients with the clinical appearance of involutional diabetic reti nopathy. Patients who had previously reFrom the Departments of Ophthalmology (Drs. W. J. Ramsay, R. C. Ramsay, and Knobloch) and Physiology (Dr. Purple), University of Minnesota, Minneapolis, Minnesota. This study was supported in part by research grants EY00293 and EY02316-01 from the National Eye Institute and the Minnesota Lions Club. This study was presented in part before the Association for Research in Vision and Ophthal mology Annual Meeting, Sarasota, Florida, April 26, 1977. Reprint requests to William H. Knobloch, M.D., Box 493 Mayo Bldg., Department of Ophthalmology, University of Minnesota, Minneapolis, ΜΝ 55455.
Fig. 1 (Ramsay and associates). Posterior pole photograph of involutional retinopathy showing optic nerve pallor, residual avascular gliosis, arteriolar and venous attenuation, pigmentary mottling, and contracture of the internal limiting membrane of the retina.
ceived extensive photocoagulation were excluded from this study. All patients had previously documented neovasculariza tion or vitreous hemorrhage and, at the time of this study, had clear media and no evidence of active proliferative disease. The patients ranged in age from 30 to 66 years (mean, 44 years), and all had insulin-dependent diabetes mellitus. The duration of diabetes ranged from 18 to 40 years (mean, 27 years). Five eyes had severe visual loss (hand movements or no light perception) secondary to traction retinal detachments and were not studied. The following determinations were made on the remaining 19 eyes: 1. Best corrected visual acuity using the methods developed by the Diabetic Retinopathy Study protocol.4 2. Goldmann visual field.
851
852
AMERICAN JOURNAL OF OPHTHALMOLOGY
3. Fundus photography of the seven standard fields identified in the Diabet ic Retinopathy Study protocol. 4 4. Dark adaptation curve using the Goldmann-Weekers dark adaptometer. 5. Fluorescein angiography. 6. Color vision using the Farnsworth-Dichotomous test (Panel D-15), the Farnsworth-Munsell 100-Hue test, and the Nagel anomaloscope. 7. Electroretinography (ERG) was performed after pupil dilatation, and initial testing was done after light adap tation at greater than 5 foot-lamberts. A Burian-Allen bipolar contact lens was used with a Grass PS-22 photostimulator placed 12 inches above the patient's eye. The stimulus measured approxi mately 30 degrees of visual field. Photopic recordings were made by using a series of photopically balanced white, red, and blue lights. Providing there was no history of a seizure disorder, the 30-Hz photopic flicker test was done to determine the phase responses to white light at maximum intensity (16 on PS 22, with no filters). In our previous control studies, we used the PS-22 photostimulator with 30 degrees of reti na illuminated and confirmed Berson and associates' 5 - 7 results regarding phase relationships obtained with a ganzfeld stimulus in the presence of a progressive tapetoretinal degeneration. Our patients were then dark-adapted for 20 minutes and the scotopic re sponses to white light flashes were re corded. Starting with the threshold in tensity necessary to evoke a measurable scotopic b-wave (10 to 20 μλθ, the in tensity was increased in steps of 0.5 log units or less in order to define the maxi mal b-wave. As b-wave amplitude was measured from the ERG baseline to the peak, care was taken in the critical area (usually between 2.0 and 2.5 log units of light attenuation) by proceeding in
DECEMBER, 1977
0.1 or 0.2 log unit increments to deter mine that point at which the scotopic a-wave began to diminish the peak b wave amplitude. After completing the scotopic series of recordings, we used deep red room light to allow transfer of the contact lens to the fellow eye. After an addition al three minutes of dark adaptation, the scotopic series was repeated. If neces sary, the patient was then light-adapted for ten minutes and the photopic series was repeated. Each eye was tested sep arately as our patients tolerated separate testing better than testing of both eyes simultaneously. Using Polaroid photographs of the oscilloscope tracings, we made meas urements from waves recorded. A-wave amplitude was measured from the ERG baseline to the trough. B-wave ampli tude was measured from the ERG base line to the peak. The latency to response onset (measured on 5 or 10 msec per division of sweep speed) and implicit time to the peak of the response were also determined. 8. Electro-oculography (EOG) was performed by recording each minute the potentials produced by ocular rota tions of 60 degrees for 15 minutes dur ing dark adaptation and then for 15 minutes during light adaptation. The light-peak (L) to dark-trough (D) ratio was then determined. The lower limit of normal for the L/D ratio in our labo ratory is 1.60. The EOG curve for each patient was plotted as microvolts per degree of ocular rotation vs time. RESULTS
The best corrected visual acuity ob tained was 6/6 (20/20) in eight eyes, 6/9 (20/30) in five eyes, 6/12 (20/40) in three eyes, 6/15 (20/50) in one eye, 6/30 (20/ 100) in one eye, and 6/60(20/200) in one eye. One eye was aphakic and all other
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
eyes had minimal refractive errors withthe spherical equivalents ranging from +0.50 to - 3 . 0 0 diopters. The visual field abnormalities recorded can be divided into two groups. One group of seven eyes showed mild abnor malities consisting of mild to moderate constriction of all isopters tested. The remaining 12 eyes had more severe abnor malities and fell into the second group. Each of these eyes had large, dense, nerve-fiber defects. Associated defects found included isolated scotomata in four eyes and a partial ring scotoma in an additional four eyes. The single eye with visual acuity of 6/60 (20/200) had the most severe visual field abnormality con sisting of a dense ring scotoma with a remaining central field of less than 10 degrees to the IV/4 target. A dark adaptation curve was deter mined in 14 eyes. Four eyes showed a delayed rod-cone break, as well as an abnormal final rod threshold which re mained elevated above a log intensity of 10 4 . The remaining ten eyes showed a delay in the rod-cone break but reached normal final rod threshold levels. The macular angiographie findings were similar in all eyes studied (Fig. 2). The abnormalities included delayed ar terial flow, areas of capillary obliteration, and minimal intraretinal vascular anoma lies. Additionally, thinning and irregular ity of the retinal pigment epithelium pro duced a mottled fluorescein pattern with window defects. No eye showed active proliferative disease, although avascular glial tissue was present along the superior and inferior vascular arcades in several eyes. The E O G was recorded in 16 eyes (Fig. 3) and all showed subnormal recordings with the L/D ratio ranging from 1.00 to 1.56 (mean, 1.33). Electroretinography was done in 17 of the 19 eyes included in this study (Fig. 4).
853
Fig. 2 (Ramsay and associates). Fluorescein angiography demonstrates delayed and incomplete arteriolar flow, capillary obliteration, and retinal pig ment epithelial atrophy with window defects.
The recordings indicated reduced a-wave and b-wave amplitudes under both photopic and scotopic testing conditions. The maximum a-wave obtained ranged from 70 μ ν to 300 μ ν with a mean of 167 μ ν . T h e maximum scotopic b-wave obtained ranged from 60 μΥ to 300 μΥ with a mean
DARK ! LIGHT
1.41
0 4
12 16 20 24 28
J I J _ J LU_ 0 4 8 12 16 20 24 28
Time in Minutes Fig. 3 (Ramsay and associates). Left, Normal curves. Right, Representative depressed curves in eyes with involutional retinopathy.
854
AMERICAN JOURNAL OF OPHTHALMOLOGY NORMAL
PROLIFERA-
IVE
INVOLUT ZONAL
PHOTOPIC 30 HZ FUCKBR
30 HZ FUCKER
SCOTOPIC
^__
1 100 miemirclts
*~* IO ora20 millistcortdi
Fig. 4 (Ramsay and associates). Photopic respons es to red, blue, and white light flashes obtained in one normal control eye and two eyes with involutional retinopathy. The response to the 30-Hz flicker (stimulus flash indicated by arrows) is abnormal in the involutional eyes. Scotopic responses to variable intensity white light flashes show marked reduction in the maximum b-wave amplitude in the involu tional eyes.
DECEMBER, 1977
of 130 μν. Normal values in our laborato ry are 200 to 500 μν for both the peak scotopic a-wave and the peak scotopic b-wave. Additionally, the latency of response to white light under scotopic conditions was delayed in all 17 eyes tested. As well, the photopic 30-Hz flicker test showed an abnormal phase relation and subnormal amplitudes in all eyes tested. Finally, b-wave oscillations were observable, but depressed in 15 eyes and absent to the highest light intensity flash in two eyes. Abnormalities of color vision were found in eight of the nine patients tested. Six patients showed evidence of an ac quired dyschromatopsia (Fig. 5) with major involvement of the tritan axis. The Famsworth-Munsell 100 Hue test was the most diagnostic. All six patients showed pathologic total error scores ranging from 145 to 471 and located around the blueyellow axis.
Fig. 5 (Ramsay and associates). In four patients. Results of Fams worth-Munsell 100 Hue tests indi cated an acquired blue-yellow dys chromatopsia.
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
855
TEST-
' ,ϊ RETEST-
I
i 12 M
Fig. 6 (Ramsay and associates). Results of Farnsworth-Munsell 100 Hue test (two trials averaged) and D-15 Panel test in one patient with involutional retinopathy are consistent with diagnosis of acquired tritanopia.
The remaining two patients showed abnormalities consistent with a diagnosis of tritanopia or tritanomaly (Fig. 6). In both cases the Raleigh Equation was within the normal range. The D-15 Panel test was normal in one patient and showed dichromatic tritanopia in the second pa tient. The Farnsworth-Munsell 100 Hue test was abnormal for both and showed a marked tritan axis with few errors scat tered outside the tritan axis. DISCUSSION
We have developed our understanding of the natural course of proliferative dia betic retinopathy through several clinical studies. 1, 8~12 Once diabetic retinal neovascularization is present, a cycle devel ops characterized by the initial growth of
the new vessels, followed by regression. Dobree 1 refers to this cycle as the primary changes in proliferative diabetic retinop athy. The final visual outcome is modi fied by secondary changes occurring in the disk, retina, vitreous, choroid, and the anterior segment of the eye, resulting from hemorrhage or connective tissue for mation in the intraocular tissues.2 The present concept of the pathogenesis of diabetic retinopathy, both preproliferative and proliferative, is based on gradually increasing retinal ischemia. 13 ' 14 While the initiating factor or factors re main unclear, capillary obliteration, shunt vessel formation, and microaneurysm formation adjacent to avascular areas represent the early histopathologic findings.16"21 Increasing retinal ischemia
856
AMERICAN JOURNAL OF OPHTHALMOLOGY
at some critical point is thought to result in neovascularization, perhaps through the elaboration of a vasoproliferative fac tor. The late regression of neovasculariza tion probably results from further vascu lar insufficiency, leading to ischemia too severe to stimulate or support vasoproliferation. Spontaneous regression of neovascular ization in the diabetic eye results in im provement of visual prognosis. The eye is generally protected from catastrophic vit reous hemorrhage and visual acuity stabi lizes, frequently at a socially useful level (vision ä 6/15 [20/50]), Despite these beneficial effects, this study clearly shows that severe functional disturbances exist in such eyes. The marked reduction in the maximum b-wave amplitude of the ERG and the depression of the EOG indicate widespread cellular abnormalities affect ing the outer layers of the retina and the retinal pigment epithelium. Additionally, all eyes studied had de layed implicit times and an abnormal phase relation to the 30-Hz photopic flicker test. As these abnormalities are characteristic of tapetoretinal degenera tions, 6 ' 7 they suggest that progressive de terioration of retinal function occurs in eyes with involutional retinopathy. While the rate of deterioration has not yet been determined, the similarity to the ERG in dominantly inherited retinitis pigmentosa suggests that several decades could elapse before the ERG becomes nonrecordable and the eye blind. We have observed several patients with a history of diabetes mellitus of 35 to 45 years duration, ad vanced involutional retinopathy, and optic nerve atrophy, with reduction of visual acuity to the level of light percep tion. Furthermore, the variation in both the depression of the ERG b-wave ampli tude and the visual field abnormalities in the eyes studied support the concept that the involutional phase of proliferative di abetic retinopathy is not an end-stage, but represents a continuum of slow deteriora
DECEMBER, 1977
tion of retinal function that may culmi nate in blindness. In this study, the extent of cellular loss, measured by the reduction of the ERG b-wave amplitude, is similar to that achieved therapeutically following panretinal photocoagulation. 22 Additional ly, when extensive photocoagulation is effective in causing regression of neo vascularization the resultant clinical ap pearance is similar to the spontaneously occurring involutional phase of diabetic retinopathy with vascular attenuation and optic nerve pallor. The ability of panretinal photocoagulation to reduce signif icantly the incidence of severe visual loss in eyes with advanced proliferative dia betic retinopathy has been demonstrated by the Cooperative Diabetic Retinopathy Study. 23 However, while Frank 22 has shown no further deterioration of the ERG response up to six weeks after exten sive photocoagulation as compared to the response immediately after treatment, the long-term effects have not been studied. If further deterioration of retinal function will occur with time, as this study sug gests, it may be necessary to evaluate the long-term effect of extensive photocoagu lation treatment. We did not expect the characteristic and repeatable tritan color vision abnormality found in eight of nine patients studied. We suppose this acquired blue-yellow dyschromatopsia is associated with the optic nerve atrophy present in all cases, rather than with a specific photoreceptor abnormality, although the findings do not distinguish between these two possibili ties. The delimited tritan defects are more characteristic of a congenital color defi ciency than of acquired defects, which generally show widespread error scores in the red-green as well as the blue-yellow axes. 24 · 25 While most patients in this study noted slow dark adaptation clinically, we could document this in only four of the 14 eyes tested. Thus, this test of rod function
VOL. 84, NO. 6
INVOLUTIONAL DIABETIC RETINOPATHY
seems affected later in the course of invol-
utional retinopathy than in that of the tapetoretinal degenerations. These results indicate that, despite ex cellent central visual acuity (17 of 19 eyes â 6/15 (20/50) and improved clinical ap pearance, eyes with involutional retinop athy show marked functional abnormali ties. We have demonstrated involvement of the retina, retinal pigment epithelium, and optic nerve; the atrophie appearance of the fundus suggests involvement of the choroidal vasculature as well. The results support the concept that involutional reti nopathy represents a continuation of the diabetic ischemie process, rather than a phase of spontaneous improvement from the earlier active proliferative phase of the disease. The data presented here suggest that the involutional phase of proliferative di abetic retinopathy may result in a pro gressive loss of retinal function. Evi dence of progression has not yet been documented. To test this hypothesis, a lon gitudinal study has been devised at the University of Minnesota. Information con cerning progressive loss of retinal func tion, as well as the rate of deterioration, is important as renal transplantation, 26 hemodialysis, and other therapeutic modali ties are used to improve the long-term survival of patients developing the microvascular complications of diabetes melli tus. SUMMARY
The end-stage or involutional phase of proliferative diabetic retinopathy may re sult in stabilization of vision for long periods of time. However, the clinical resemblance to the progressive tapetore tinal degenerations suggests that marked functional impairment of the retina is present in such eyes. We studied 19 eyes with involutional retinopathy to document the status of the retinal function. Studies included fluo
857
rescein angiography, visual field exami nation, dark adaptation testing, color vi sion testing, electro-oculography and electroretinography (ERG). The results indicated marked function al abnormalities in all eyes. The ERG tracings showed uniformly subnormal re sponses and delayed implicit times, simi lar to those of dominantly inherited reti nal pigment degeneration, and indicative of a progressive retinal disorder. In two patients, color vision testing showed de fects similar to those seen in inherited tritanopia; and in the remaining patients, defects were indicative of an acquired blue-yellow dyschromatopsia. REFERENCES 1. Dobree, J. H.: Proliferative diabetic retinopa thy. Evolution of the retinal lesions. Br. J. Ophthalmol. 48:637, 1964. 2. Davis, M. D.: Vitreous contraction in prolifera tive diabetic retinopathy. Arch. Ophthalmol. 74:741, 1965. 3. Peyman, G. A., Huamonte, F., Locketz, A., and Goldberg, M. F.: Fluorescein angiography of the fundus after pars plana vitrectomy. Ann. Oph thalmol. 8:791, 1976. 4. Diabetic Retinopathy Study: Manual of Opera tions. Baltimore, Diabetic Retinopathy Study Coor dinating Center, 1972. 5. Berson, E. L., Gouras, P., and Hoff, M.: Tem poral aspects of the electroretinogram. Arch. Oph thalmol. 81:207, 1969. 6. Berson, E. L., Gouras, P., and Gunkel, R. D.: Rod responses in retinitis pigmentosa, dominantly inherited. Arch. Ophthalmol. 80:58, 1968. 7. : Progressive cone-rod degeneration. Arch. Ophthalmol. 80:68, 1968. 8. Burditt, A. F., Caird, F. I., and Draper, G. J.: The natural history of diabetic retinopathy. Q. J. Med. 37:303, 1968. 9. Davis, M. D.: The natural course of diabetic retinopathy. Trans. Am. Acad. Ophthalmol. Otolaryngol. 72:237, 1968. 10. Beethan, W. P.: Visual prognosis of proliferat ing diabetic retinopathy. Br. J. Ophthalmol. 47:611, 1963. 11. Davis, M.: Natural course of diabetic retinop athy. In Kimura, S. J., and Caygill, W. M. (eds.): Vascular Complications of Diabetes Mellitus with Special Emphasis on Microangiography of the Eye. St. Louis, C. V. Mosby, 1967, p. 139-169. 12. Dobree, J. H.: Evolution of lesions in prolife rative diabetic retinopathy. An 8-year photographic survey. In Goldberg, M., and Fine, S. (eds.): The Treatment of Diabetic Retinopathy. Washington, U.S. Public Health Service, 1969, p. 55-64.
858
AMERICAN JOURNAL OF OPHTHALMOLOGY
13. Davis, M., Myers, F., Engerman, R., DeVenecia, G., and Magli, Y.: Clinical observations con cerning the pathogenesis of dgabetic retinopathy. In Goldberg, M., and Fine, S. (eds.): The Treatment of Diabetic Retinopathy. Washington, U.S. Public Health Service, 1969, p. 47-53. 14. Goldberg, M. F.: The role of ischemia in the production of vascular retinopathies. In Lynn, J., Snyder, W., and Vaiser, A. (eds.): Diabetic Retinopa thy. New York,Gruneand Stratton, 1974, p. 47-63. 15. Cogan, D. G., and Kuwabara, T.; Capillary shunts in the pathogenesis of diabetes retinopathy. Diabetes 12:293, 1963. 16. Ashton, N.: Arteriolar involvement in diabetic retionpathy. Br. J. Ophthalmol. 37:282, 1953. 17. : Vascular basement membrane chang es in diabetic retinopathy. Br. J. Ophthalmol. 58: 344, 1974. 18. Ditzel, J.: Haemorheological factors in the development of diabetic microangiopathy. Br. J. Ophthalmol. 51:793, 1967. 19. Cogan, D. G., Troussaint, D., and Kuwabara, T.: Retinal vascular patterns. Arch. Ophthalmol. 66:100, 1961.
DECEMBER, 1977
20. Kuwabara, T., and Cogan, D. G.: Retinal vascular patterns. 6. Mural cells of the retinal capil laries: Arch. Ophthalmol. 69:114, 1963. 21. Yanoff, M.: Ocular pathology of diabetes mellitus. Am. J. Ophthalmol. 67:21, 1969. 22. Frank, N.: Visual fields and electroretinography following extensive photocoagulation. Arch. Ophthalmol. 93:591,1975. 23. Diabetic Retinopathy Study Research Group: Preliminary report on effects of photocoagulation therapy. Am. J. Ophthalmol. 81:383, 1976. 24. Krill, A. E., Smith, V. C , and Pokorny, J.: Further studies supporting the identity of congeni tal tritanopia and hereditary dominant optic atro phy. Invest. Ophthalmol. 10:182,1971. 25. —Similarities between congenital tritan de fects and dominate optic nerve atrophy. Coinci dence or identity. J. Optom. Soc. Am. 60:1132, 1970. 26. Najarian, J. S., Sutherland, D. E., Simmons, R. L., Howard, R. J., Kjellstrand, C. M., Mauer, S. M., Kennedy, W., Ramsay, R. C , Barbosa, J., and Goetz, F. C : Kidney transplantation for the uremie diabet ic. Surg. Gynecol. Obstet. 144:682, 1977.