- Email: [email protected]
Macular Phototoxicity Caused by Fiberoptic Endoillumination During Pars Plana Vitrectomy
Macular Phototoxicity Caused by Fiberoptic Endoillumination During Pars Plana Vitrectomy Mark M i c h e l s , M . D . , H i l e l L e w i s , M . D . , Gary W. A b r a m s , M . D . , D e n n i s P. Han, M . D . , W i l l i a m F. M i e l e r , M . D . , and Jay N e i t z , P h . D .
Three cases of retinal phototoxicity were caused by light from the endoilluminator used in vitrectomy. Preoperative clinical examination, color photography, and fluorescein angiography in all three cases failed to disclose retinal pigment epithelial changes. Postoperative clinical findings, photographs, and fluorescein angiograms are highly suggestive of the presence of retinal phototoxicity. The characteristics of these lesions and surgical conditions implicate the endoilluminator as the source of photic injury. The macular l e sions were noted within one week of the surgical procedure, measured between 2 and 5 disk diameters in size, involved the fovea in two eyes, and resulted in marked decrease in visual acuity in two of three eyes, with persistence in one eye. Initially, whitening of the outer retina was present, but was replaced by pigmentary mottling at the level of the retinal pigment epithelium within a few weeks. Preventive measures to avoid macular phototoxicity associated with vitrectomy are discussed. attributed to the oper ating microscope has been described after many ophthalmic surgical procedures including vit-
R E T I N A L PHOTOTOXICITY
Accepted for publication June 5, 1992. From the Retina Division, Jules Stein Eye Institute and Department of Ophthalmology, UCLA School of Medi cine, Los Angeles, California (Drs. Michels and Lewis); and the Eye Institute, Medical College of Wisconsin, Milwaukee, Wisconsin (Drs. Abrams, Han, Mieler, and Neitz). This study was supported in part by a manpower award from Research to Prevent Blindness, Inc., New York, New York (Dr. Lewis); by the Charles Kenneth Feldman Fund, Los Angeles, California (Dr. Lewis); and by an unrestricted grant from Research to Prevent Blind ness, Inc., New York, New York, to the Eye Institute, Medical College of Wisconsin, Milwaukee, Wisconsin. Reprint requests to Hilel Lewis, M.D., Jules Stein Eye Institute, 100 Stein Plaza, Los Angeles, CA 90024-7007.
rectomy, 1 and has been extensively re viewed. 2 The first case attributed to the intra ocular light source used in vitrectomy was described after epiretinal membrane dissection and was attributed to prolonged exposure to direct illumination during dissection of partic ularly adherent epicenters. 3 One additional case of endoilluminator-related retinal photo toxicity has been described in humans and also occurred during dissection of a difficult epireti nal membrane. 4 Mottling of the retinal pigment epithelium has been described after a macular hole surgical procedure and, although not ad dressed in pre- or postoperative photographs, has been attributed to the light of the endoillu minator. 6 To these cases we add three addition al cases of macular phototoxicity caused by the fiberoptic endoilluminator. We studied the shared features of these three cases and preven tive measures.
Patients and Methods During the two-year period between 1989 and 1991, four of us (H.L., G.W.A., D.P.H., and W.F.M.) operated on 132 eyes (132 patients) with epiretinal membranes in the macular re gion and recognized three eyes (three patients, 2.3%) that developed postoperative macular lesions consistent with retinal phototoxicity. All three patients were examined pre- and postoperatively and were operated on by one of us at the Jules Stein Eye Institute or at the Medical College of Wisconsin. A complete ocular histo ry and examination were performed in each case. Preoperative and postoperative examina tions included contact-lens fundus biomicroscopy, indirect ophthalmoscopy, and fluorescein angiography. The surgical procedure included a three-port pars plana vitrectomy and removal of the epiretinal membrane by using a barbed microvitreoretinal blade and intraocular for-
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:287-296, SEPTEMBER, 1992
287
288
AMERICAN JOURNAL OF OPHTHALMOLOGY
September, 1992
ceps. The Grieshaber (Langhorne, Pennsylva nia) or Trek (Mukwonago, Wisconsin) light source using maximum illumination was used. The cornea was covered when no intraocular operation was being performed to minimize retinal exposure to the operating microscope light in all cases.
Case Reports Casel A 43-year-old woman was referred to us in December 1991 with a macular pucker in her left eye. Her ocular and medical history were remarkable for bilateral acute retinal necrosis in 1981, and multiple sclerosis in 1986. The right eye developed a retinal detachment and after failed surgical therapy, the eye was enu cleated in 1983. The left eye developed neovascularization at the optic nerve head for which panretinal photocoagulation was performed in 1981. Between 1981 and 1986, the patient had transient neurologic symptoms including inter mittent decreases in vision that lasted approxi mately one week, but were unassociated with retrobulbar pain on eye movement. Magnetic resonance imaging demonstrated findings typi cal of multiple sclerosis, which was thought to account for her symptoms. Since 1986, she complained of intermittent disturbances of vi sion approximately three to four times per year. With systemic prednisone, which she had taken since 1989, the symptoms resolved. Until 1989, the visual acuity in the left eye could be corrected to 2 0 / 2 5 . Since then, she noted a slowly progressive decrease in visual acuity and metamorphopsia in the left eye. In 1990, she was unable to drive and required low-vision aids to read. Our first examination on Dec. 5, 1990, disclosed a best-corrected visual acuity of 20/125. Anterior segment ex amination snowed no signs of active inflamma tion and no neovascularization of the iris. The lens and vitreous were clear. The optic nerve head showed mild temporal pallor and was sharp. In the macular region, there was an epiretinal membrane with a pseudohole and mild retinal vascular tortuosity (Fig. 1). Periph eral photocoagulation scars were noted posteri or to the area of previous retinal necrosis. Fluorescein angiography demonstrated no reti nal pigment epithelial hyperfluorescence or marked macular edema (Fig. 2). After discuss ing the alternatives with the patient, she decid-
Fig. 1 (Michels and associates). Case 1. Preopera tive photograph of the left eye shows an epimacular membrane. Visual acuity was 20/125. ed to undergo pars plana vitrectomy and mem brane dissection, which was performed one week later. Intraoperatively, the epiretinal membrane was friable, adhered to the retina, and difficult to remove. The intraocular endoilluminator and light source were used. External illumination was provided by an operating mi croscope fitted with a 100-W halogen bulb transmitted with a fiberoptic cable. Total surgi cal time was one hour 50 minutes. Skin temper ature varied from 33.8 to 34.5 C. Oxygen satu ration was 99%. On the first postoperative day, her visual acuity was hand movements. By indirect ophthalmoscopy and biomicroscopy, a creamy oval area at the level of the deep retina and involving the fovea was noted. This area measured approximately 4 x 5 disk diameters. Photographs taken one week postoperatively demonstrated that this discrete lesion primarily involved the retinal pigment epithelium (Fig. 3). Fluorescein angiography performed two weeks postoperatively demonstrated discrete early mottled hyperfluorescence with late stain ing in the affected area (Fig. 4, left). Late stain ing of the overlying retinal vein could also be seen (Fig. 4, right). In March 1991, the visual acuity improved to 20/300 and the macular lesion showed mottled pigmentation. The pa tient was last examined in October 1991, at which time the visual acuity remained at 2 0 / 300.
Vol. 114, No. 3
Macular Phototoxicity
Fig. 2 (Michels and associates). Case 1. Preopera tive fluorescein angiogram shows normal retinal and choroidal vascular filling. Case 2 A 76-year-old w o m a n w h o h a d a five-month history of b l u r r e d vision in h e r r i g h t eye w a s e x a m i n e d in April 1 9 9 1 . H e r b e s t - c o r r e c t e d visual acuity w a s R.E.: 2 0 / 2 0 0 a n d L.E.: 2 0 / 2 0 . The anterior segment was normal. Examination of the r i g h t eye w i t h b i o m i c r o s c o p y a n d i n d i -
289
Fig. 3 (Michels and associates). Case 1. One week after vitrectomy, a creamy oval discrete lesion involv ing the outer retina and retinal pigment epithelium was seen in the macula region. Visual acuity was hand motions. rect o p h t h a l m o s c o p y s h o w e d a p r o m i n e n t m a c u l a r e p i r e t i n a l m e m b r a n e c a u s i n g fractional e l e v a t i o n of t h e fovea (Fig. 5). T h e r e w e r e p r o m i n e n t r e t i n a l striae a n d r e t i n a l m i c r o v a s cular a b n o r m a l i t i e s o n t h e s u p e r i o r e d g e of t h e
Fig. 4 (Michels and associates). Case 1. Postoperative fluorescein angiogram. Left, Early phase shows a discrete lesion with mottled fluorescence and hyperfluorescent margins. Right, Venous phase shows increased fluores cence and staining of the retinal vein overlying the lesion and at the level of the sclera.
290
AMERICAN JOURNAL OF OPHTHALMOLOGY
September, 1992
Fig. 5 (Michels and associates). Case 2. Prominent epiretinal membrane causing a tractional detachment of the fovea. Visual acuity was 20/200.
Fig. 6 (Michels and associates). Case 2. Three weeks postoperatively, the right eye demonstrated pigment clumping in the papillomacular bundle and surrounding the fovea. Visual acuity was 20/400.
epiretinal membrane. Pigmentary changes were not present. Fluorescein angiography also failed to show abnormalities at the level of the retinal pigment epithelium. On April 16, 1991, she underwent a three-port pars plana vitreetomy and membrane dissection in the right eye.
A thick and linear epiretinal membrane that extended across the macula was easily removed. Another epiretinal membrane on the inferonasal edge of the macula was also removed. The traction on the macula was released and no complications occurred. At the
Fig. 7 (Michels and associates). Case 2. Postoperative fluorescein angiogram. Left, Arteriovenous phase demonstrated mottled fluorescence of the lesion surrounding the fovea and in the papillomacular bundle. Right, Late phase of the angiogram shows staining of the area corresponding to the pigmentary disturbance noted clinically.
Vol. 114, No. 3
Macular Phototoxicity
end of the surgical procedure, cryotherapy was used to treat two preexisting anterior retinal tears, and a fluid-air exchange was performed. Total surgical time was one hour 50 minutes, and the intraocular endoilluminator and light source were used. External illumination was provided by a microscope fitted with a 50-W tungsten filament bulb. The patient's intraoper ative core temperature was 35.0 to 35.3 C. The oxygen saturation was 99%. One week postoperatively, visual acuity in the right eye was 20/400 and retinal examina tion disclosed macular edema. Three weeks postoperatively, hypopigmented mottling throughout the nasal portion of the macula and surrounding the fovea was noted at the level of the retinal pigment epithelium (Fig. 6). Fluores cein angiography showed early hyperfluorescence and late staining in the area where the pigmentary changes were seen clinically (Fig. 7). Four months postoperatively, the visual acu ity in her right eye had improved to 2 0 / 6 0 - 2 , although the macular pigmentary changes re mained unchanged. Case 3 A 72-year-old man with a visual acuity of 1/200 in the right eye underwent a pneumatic retinopexy for a macula-off rhegmatogenous retinal detachment on Aug. 26, 1988. On Sept. 29, 1988, visual acuity had improved to 20/80
Fig. 9 (Michels and associates). Case 3. Left, One ir demonstrated a persistent epimacular membrane. Vi shows persistent macular edema and no abnormalities
291
Fig. 8 (Michels and associates). Case 3. Four months after pneumatic retinopexy, an epimacular membrane was noted. Visual acuity was 20/200.
and the retina was totally attached. On Dec. 28, 1988, the visual acuity in the right eye de creased to 20/200 because of the development of an epiretinal membrane in the macular re gion (Fig. 8). On March 3, 1989, a pars plana vitrectomy and epiretinal membrane dissection
ith after vitrectomy for macular pucker, the right eye al acuity was 20/100. Right, Fluorescein angiogram the level of the retinal pigment epithelium or choroid.
292
AMERICAN JOURNAL OF OPHTHALMOLOGY
in the right eye were performed. On April 6, 1989, the visual acuity in the right eye was 20/100 and a persistent epiretinal membrane on the superotemporal macula that exerted traction on the fovea was noted (Fig. 9). On April 14, 1989, a repeat pars plana vitrectomy with epiretinal membrane dissection was per formed. The epiretinal membrane adhered to the retina and fragmented while it was being removed. No intraoperative complications oc curred. The Trek intraocular endoilluminator and light source were used. External illumina tion was provided by an operating microscope
September, 1992
fitted with a 100-W halogen bulb transmitted with a fiberoptic cable. Total surgical time was two hours 25 minutes. The oxygen saturation was 99 % andcore temperature was 35.5 to 36.4 C. Five days postoperatively, the visual acuity in the right eye was 20/70. The macula showed mild white opacification and edema of the outer retina involving the superior macula. On May 18, 1989, the visual acuity in the right eye was 20/80. Fundus examination disclosed a sharply demarcated area of hypopigmentation at the level of the pigment epithelium in the supero nasal macula (Fig. 10). Fluorescein angiogra-
Fig. 10 (Michels and associates). Case 3. Top left, One month after repeat vitrectomy for a persistent epimacular membrane, an area of hypopigmentation appeared superonasal to the fovea at the level of the retinal pigment epithelium. Visual acuity was 20/80. Top right, Early arteriovenous phase of the fluoresce in angiogram demonstrates minimal mottled hyperfluorescence. Bottom left, Venous phase of the fluo rescein angiogram demonstrates increased fluo rescence of the lesion superonasal to the fovea and mild macular edema.
Vol. 114, No. 3
Macular Phototoxicity
phy disclosed mottled hyper- and hypofluorescence corresponding to the area of retinal pigment epithelial hypopigmentation seen clin ically.
Results The epiretinal membranes in the macular region in our cases were secondary to a rhegmatogenous retinal detachment in one eye, a pos terior vitreous separation with the develop ment of retinal tears in one eye, and acute retinal necrosis in one eye. Preoperatively, the visual acuity ranged from 20/100 to 20/200. Intraoperatively, the epiretinal membranes ad hered to the retina and were difficult to remove in two of the three cases. In these two cases, the endoillumination was held close to the retina, and the surgical procedure was prolonged. Ret inal exposure to the light of the operating microscope, however, was limited to approxi mately 20 minutes for standard opening and closing associated with a three-port vitrectomy. In all three cases, macular lesions consisting of white opacification of the outer retina were noted within one week postoperatively. These findings were not noted intraoperatively. The macula was edematous, and in one eye (Case 1) acute segmental retinal phlebitis was also pres ent overlying the area of outer retinal opacifica tion. The macular lesions were all sharply de marcated, irregularly oval to round, varied in size from 2 to 5 disk diameters, and involved the fovea in two eyes. Initially, the visual acuity decreased in two eyes and improved in one eye. With time, visual acuity improved in the eyes with initial postoperative decrease in vision. The visual acuity of one eye improved to sub stantially better than the preoperative value. The time to recovery of best visual acuity varied from three to 12 months. With time, the white opacification of outer retina changed and was characterized by mottling of the retinal pig ment epithelium.
Discussion In our three cases, the characteristic course of the macular lesion and its fluorescein angio graphie features were typical of retinal photo toxicity.2 Phototoxic lesions are typically noted clinically only several days after the light expo
293
sure as in our three cases. A discrete margin of retinal pigment epithelial hypopigmentation is noted to develop as the initial outer retinal whitening fades. Fluorescein angiography clearly demonstrates the area of retinal pig ment epithelial disturbance, as early discrete hyperfluorescence throughout the area of mot tled pigmentary change evolves to late staining but never leakage. In most cases, the pigment clumping results in a monotonous, diffuse mot tled hyperfluorescence that can evolve into a target-like central clumping surrounded by hy popigmentation. Typically, lesions caused by the operating microscope assume the shape of the light source used. 6 The lesion produced by tungsten filaments often assumes the oval shape of the filament, whereas fiberoptic systems often gen erate a round lesion. 1 These lesions tend to have sharply demarcated borders partly because of minimal ocular or light position change during anterior segment procedures. The size of the lesion is usually less than 1 to 2 disk diameters. Our cases demonstrated the typical clinical and fluorescein angiographie features of opera ting microscope-induced retinal phototoxicity. The features that differentiate the lesions in our cases from the typical operating microscopeinduced lesion are the size and shape of the lesions and the minimal exposure to light from the operating microscope. In all three cases, the lesions were markedly larger than would be expected with the operating microscope light source. In Case 2, a nearly round lesion was seen despite the use of a tungsten filament in the microscope light source, which should re sult in a lesion of much smaller and more rectangular dimensions. The lesions in our cases had marked pleomorphism, probably be cause of intraoperative alterations in illumina tion positioning in terms of proximity to the retinal surface as well as location of illumina tion, which is typical of surface-level vitrec tomy, especially when it involves dissection of epiretinal membranes over a large area. The lesion in Case 3, however, had a fairly discrete epiretinal membrane above the fovea. Surface dissection involved a discrete area and the light exposure was likely limited largely to this area. This probably accounts for the comparatively discrete appearance of the phototoxic lesion. Similarly, anterior segment procedures which may result in operating microscope-induced retinal phototoxicity and experimental endoilluminator-induced retinal phototoxicity are characterized by minimal change in light
294
September, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
position and therefore may result in discrete lesions. 7 Retinal phototoxicity is a function of the following three factors: wavelength of the light source, duration of exposure, and power level. Duration of exposure and power level deter mine total energy delivered to a target tis sue. 8 For retinal phototoxicity to occur, energy must be transmitted to and absorbed by a target tissue. With an operating microscope, light is transmitted through the whole globe, a process that has been carefully studied. In the intact globe, reflection by and absorption in the tis sues anterior to the retina protects the retina from the short-wavelength ultraviolet light and some of the blue visible light that is most efficient in inducing phototoxicity. 910 With an endoilluminator, there is no absorp tion of damaging irradiation by anterior ocular structures once the vitreous is cleared of opaci ty. Fuller, Machemer, and Knighton 7 found that clinical phototoxicity can be seen after a 15minute exposure and histologie damage can be noted after just a 10-minute exposure in owl monkeys. In their carefully controlled study, monkeys were exposed to between five and 30 minutes of endoillumination, and then studied by ophthalmoscopy, fluorescein angiography, and light and electron microscopy at several intervals after exposure. As is true of all cases of photochemical damage, lesions were not immediately apparent even after the 30-minute exposure, but could be seen as early as five hours after exposure. This observation is criti cal for the vitreoretinal surgeon who is accus tomed to observing immediate thermal effects delivered by other forms of electromagnetic irradiation, such as the endolaser. 8 Kuhn, Morris, and Massey 4 suggested that the danger of iatrogenic photic damage from the endoilluminator probe with subsequent visual impairment is greatest when vitrectomy is per formed for epiretinal membrane dissection. They concluded that the problem is not widely recognized because the subtle changes clearly seen by fluorescein angiography are likely to be overlooked on ophthalmoscopy. We further suggest that many of the characteristic angio graphie findings noted at the level of the retinal pigment epithelium may not be present when retinal phototoxicity occurs in detached retina, as has been experimentally demonstrated by Zilis and Machemer. 11 In this instance, damage to noncontiguous retinal pigment epithelium may not be seen even with fluorescein angiog raphy.
Protection of the retina from endoilluminator-associated retinal phototoxicity is provided in part by macular xanthophyll. Jaffe and Wood12 were the most recent to demonstrate this qualitatively. Our Case 2 demonstrates macular sparing that may be explained in part by the macular xanthophyll. Ham and associ ates 8 demonstrated a twofold threshold energy increase required to produce a foveal compared to a parafoveal lesion in monkeys. Although no direct evidence links wave length attenuation with reduction of endoilluminator-associated retinal phototoxicity, ex perimental work in phakic dogs 13 and rhesus monkeys' 4 suggests that limitation of blue light exposure markedly reduces the incidence of indirect ophthalmoscope-induced and of oper ating microscope-induced retinal phototoxi city, respectively. Although Fuller, Machemer, and Knighton 7 documented less than 5% blue light output from their light pipe endoilluminator, and Meyers and Bonner' 5 measured that the blue fraction is less than 3 % of transmitted energy, both groups of authors emphasized that elimination of light less than 500 nm from fiberoptic endoilluminators may decrease the risk of retinal phototoxicity by tenfold. Experimental proof of the benefit of blue light filtration in the endoilluminator probe is lacking. Other factors that are somewhat controllable are the power level of the light source and the exposure time. The currently popular cooler (and bluer) light sources, such as the xenon ~ 20 CM
E E w 15 CO
5 = 10
î
0
«-
10
15
20
25
Distance From Endoilluminator tip (mm)
Fig. 11 (Michels and associates). Endoilluminator irradiance as a function of the distance from the light source (mW/mm2) in air measured using the Trek endoilluminator and light source.
Vol. 114, No. 3
Macular Phototoxicity
source, and the higher intensity often needed for video recording equipment need to be con sidered when performing a surgical procedure. Maximum output of the Trek endoilluminator and light source and the Grieshaber endoillu minator and light source was determined to be 15.6 and 18.0 mW/mm 2 , respectively, using a silicone diode photodetector (UDT model 247) traceable to the National Bureau of Standards. Although power level is not often considered a flexible variable, it is noteworthy that the reti nal irradiance value is inversely proportional to the square of the distance of the source light to the retinal surface.16 Thus, moving the light source as far away from the retinal surface as is possible during noncritical periods of mem brane dissection seems prudent (Fig. 11). With regard to exposure time, some investigators have advocated intermittent exposure as less likely to cause the same degree of damage as constant exposure for a given cumulative time period. 4 The importance of intermittent expo sure is debatable. Data on brief intermittent exposures in pseudophakic primates 17 and phakic primates 18 contend that these exposures are cumulative and that intermittent intraoper ative rest periods are unlikely to be protective. There are two variables independent of the immediate ocular milieu that probably do have a role in retinal phototoxicity. System tempera ture has long been recognized as important in retinal phototoxicity. The idea that the photo chemical process might be enhanced by in creased core temperature was addressed and shown by Noell and associates. 19 Rinkoff and associates 20 demonstrated that cooling infusion fluid from 39 C to 29 C in rabbits in which vitrectomy had been performed while closely illuminating the retina with an endoilluminator increased the damage threshold time from 25 to 60 minutes. Although neither alteration of core temperature nor use of hypothermie infusion fluid has been studied in humans, it is prudent to avoid unnecessary increases in core or infu sion fluid temperature. The other variable not immediately associat ed with the globe is tissue oxygénation. Be cause the photochemical process is thought to result from the action of free radical and singlet oxygen moieties, high levels of inspired oxygen might decrease the threshold for phototoxicity. Jaffe and associates 21 studied the role of in spired oxygen in phakic rhesus monkeys ex posed to operating microscope light under con ditions of 2 1 % Fi0 2 and 99% Fi0 2 and noted marked potentiation of phototoxic damage un
295
der conditions of increased oxygénation. Al though oxygénation has not been directly stud ied under conditions of endoilluminatorrelated retinal phototoxicity, it seems prudent to avoid unnecessary oxygénation in patients undergoing a surgical procedure using the en doilluminator. Endoilluminator-related retinal phototoxic ity is a largely unrecognized, and partially preventable, sequela to certain cases of vitrec tomy. Awareness of the potential for its occur rence and minimization of both intraocular and extraocular variables that may contribute to the process are advised.
References 1. McDonald, H. R., and Harris, M. J.: Operating microscope induced retinal phototoxicity during pars plana vitrectomy. Arch. Ophthalmol. 106:521, 1988. 2. Michels, M., and Sternberg, P.: Operating mi croscope-induced retinal phototoxicity. Pathophysiology, clinical manifestations and prevention. Surv. Ophthalmol. 34:237, 1990. 3. McDonald, H. R., Verre, W. P., and Aaberg, T. M.: Surgical management of idiopathic epiretinal membranes. Ophthalmology 93:978, 1986. 4. Kuhn, F., Morris, R., and Massey, M.: Photic retinal injury from endoillumination during vitrec tomy. Am. J. Ophthalmol. 111:42, 1991. 5. Kelly, N. E., and Wendel, R. T.: Vitreous sur gery for idiopathic macular holes. Results of a pilot study. Arch. Ophthalmol. 109:654, 1991. 6. Khwarg, S. G., Linstone, F. A., Daniels, S. A., Isenberg, S. J., Hanscom, T. A., Geoghegan, M., and Straatsma, B. R.: Incidence, risk factors, and mor phology in operating microscope light retinopathy. Am. J. Ophthalmol. 103:255, 1987. 7. Fuller, D., Machemer, R„ and Knighton, R. W.: Retinal damage produced by intraocular fiber optic light. Am. J. Ophthalmol. 85:519, 1978. 8. Ham, W. T„ Ruffolo, J. J., Mueller, H. A., and Guerry, D.: The quantitative dimensions of intense light damage as obtained from animal studies. The nature of retinal radiation damage. Dependence on wavelength, power level and exposure time. Vision Res. 20:1105, 1980. 9. Boettner, E. A., and Wolter, J. R.: Transmission of the ocular media. Invest. Ophthalmol. Vis. Sci. 1:776, 1962. 10. Ham, W. T., Mueller, H. A., Ruffolo, J. J., and Clarke, A. M.: Sensitivity of the retina to radiation damage as a function of wavelength. Photochem. Photobiol. 29:735, 1979. 11. Zilis, J. D., and Machemer, R.: Light damage in detached retina. Am. J. Ophthalmol. 111:47, 1991. 12. Jaffe, G. J., and Wood, I.: Retinal phototoxicity
296
AMERICAN JOURNAL OF OPHTHALMOLOGY
from the operating microscope. Protective effect of the fovea. Ophthalmology 106:445, 1988. 13. Buyukmihci, N.: Photic retinopathy in the dog. Exp. Eye Res. 33:95, 1981. 14. Weil, L., Michels, M., Sternberg, P., and Aaberg, T. M.: Filtration of blue (<515nm) light significantly reduces retinal phototoxicity. ARVO ab stracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1989, p. 459. 15. Meyers, S., and Bonner, R. F.: Retinal irradiance from vitrectomy endoilluminators. Am. J. Oph thalmol. 94:26, 1982. 16. Calkins, J. L., Hochheimer, B. F., and D'Anna, A.: Potential hazards from specific ophthalmic devic es. Vision Res. 20:1039, 1980. 17. Irvine, A. R., Wood, I., and Morris, B. W.: Reti nal damage from the illumination of the operating
September, 1992
microscope. An experimental study in pseudophakic monkeys. Arch. Ophthalmol. 102:1358, 1984. 18. Lawwill, T., Crockett, R. S., and Currier, G.: Retinal damage secondary to chronic light exposure. Thresholds and mechanisms. Doc. Ophthalmol. 44:379, 1977. 19. Noell, W. K., Walker, V. S., Kang, B. S., and Berman, S.: Retinal damage by light in rats. Invest. Ophthalmol. 5:450, 1966. 20. Rinkoff, J., Machemer, R., Hida, T., and Chan dler, D.: Temperature-dependent light damage to the retina. Am. J. Ophthalmol. 102:452, 1986. 21. Jaffe, G. J., Irvine, A. R., Wood, I. S., Severinghaus, J. W., Pino, G. R., and Haugen, C : Retinal phototoxicity from the operating microscope. The role of inspired oxygen. Ophthalmology 95:1130, 1988.
Three cases of retinal phototoxicity were caused by light from the endoilluminator used in vitrectomy. Preoperative clinical examination, color photography, and fluorescein angiography in all three cases failed to disclose retinal pigment epithelial changes. Postoperative clinical findings, photographs, and fluorescein angiograms are highly suggestive of the presence of retinal phototoxicity. The characteristics of these lesions and surgical conditions implicate the endoilluminator as the source of photic injury. The macular l e sions were noted within one week of the surgical procedure, measured between 2 and 5 disk diameters in size, involved the fovea in two eyes, and resulted in marked decrease in visual acuity in two of three eyes, with persistence in one eye. Initially, whitening of the outer retina was present, but was replaced by pigmentary mottling at the level of the retinal pigment epithelium within a few weeks. Preventive measures to avoid macular phototoxicity associated with vitrectomy are discussed. attributed to the oper ating microscope has been described after many ophthalmic surgical procedures including vit-
R E T I N A L PHOTOTOXICITY
Accepted for publication June 5, 1992. From the Retina Division, Jules Stein Eye Institute and Department of Ophthalmology, UCLA School of Medi cine, Los Angeles, California (Drs. Michels and Lewis); and the Eye Institute, Medical College of Wisconsin, Milwaukee, Wisconsin (Drs. Abrams, Han, Mieler, and Neitz). This study was supported in part by a manpower award from Research to Prevent Blindness, Inc., New York, New York (Dr. Lewis); by the Charles Kenneth Feldman Fund, Los Angeles, California (Dr. Lewis); and by an unrestricted grant from Research to Prevent Blind ness, Inc., New York, New York, to the Eye Institute, Medical College of Wisconsin, Milwaukee, Wisconsin. Reprint requests to Hilel Lewis, M.D., Jules Stein Eye Institute, 100 Stein Plaza, Los Angeles, CA 90024-7007.
rectomy, 1 and has been extensively re viewed. 2 The first case attributed to the intra ocular light source used in vitrectomy was described after epiretinal membrane dissection and was attributed to prolonged exposure to direct illumination during dissection of partic ularly adherent epicenters. 3 One additional case of endoilluminator-related retinal photo toxicity has been described in humans and also occurred during dissection of a difficult epireti nal membrane. 4 Mottling of the retinal pigment epithelium has been described after a macular hole surgical procedure and, although not ad dressed in pre- or postoperative photographs, has been attributed to the light of the endoillu minator. 6 To these cases we add three addition al cases of macular phototoxicity caused by the fiberoptic endoilluminator. We studied the shared features of these three cases and preven tive measures.
Patients and Methods During the two-year period between 1989 and 1991, four of us (H.L., G.W.A., D.P.H., and W.F.M.) operated on 132 eyes (132 patients) with epiretinal membranes in the macular re gion and recognized three eyes (three patients, 2.3%) that developed postoperative macular lesions consistent with retinal phototoxicity. All three patients were examined pre- and postoperatively and were operated on by one of us at the Jules Stein Eye Institute or at the Medical College of Wisconsin. A complete ocular histo ry and examination were performed in each case. Preoperative and postoperative examina tions included contact-lens fundus biomicroscopy, indirect ophthalmoscopy, and fluorescein angiography. The surgical procedure included a three-port pars plana vitrectomy and removal of the epiretinal membrane by using a barbed microvitreoretinal blade and intraocular for-
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:287-296, SEPTEMBER, 1992
287
288
AMERICAN JOURNAL OF OPHTHALMOLOGY
September, 1992
ceps. The Grieshaber (Langhorne, Pennsylva nia) or Trek (Mukwonago, Wisconsin) light source using maximum illumination was used. The cornea was covered when no intraocular operation was being performed to minimize retinal exposure to the operating microscope light in all cases.
Case Reports Casel A 43-year-old woman was referred to us in December 1991 with a macular pucker in her left eye. Her ocular and medical history were remarkable for bilateral acute retinal necrosis in 1981, and multiple sclerosis in 1986. The right eye developed a retinal detachment and after failed surgical therapy, the eye was enu cleated in 1983. The left eye developed neovascularization at the optic nerve head for which panretinal photocoagulation was performed in 1981. Between 1981 and 1986, the patient had transient neurologic symptoms including inter mittent decreases in vision that lasted approxi mately one week, but were unassociated with retrobulbar pain on eye movement. Magnetic resonance imaging demonstrated findings typi cal of multiple sclerosis, which was thought to account for her symptoms. Since 1986, she complained of intermittent disturbances of vi sion approximately three to four times per year. With systemic prednisone, which she had taken since 1989, the symptoms resolved. Until 1989, the visual acuity in the left eye could be corrected to 2 0 / 2 5 . Since then, she noted a slowly progressive decrease in visual acuity and metamorphopsia in the left eye. In 1990, she was unable to drive and required low-vision aids to read. Our first examination on Dec. 5, 1990, disclosed a best-corrected visual acuity of 20/125. Anterior segment ex amination snowed no signs of active inflamma tion and no neovascularization of the iris. The lens and vitreous were clear. The optic nerve head showed mild temporal pallor and was sharp. In the macular region, there was an epiretinal membrane with a pseudohole and mild retinal vascular tortuosity (Fig. 1). Periph eral photocoagulation scars were noted posteri or to the area of previous retinal necrosis. Fluorescein angiography demonstrated no reti nal pigment epithelial hyperfluorescence or marked macular edema (Fig. 2). After discuss ing the alternatives with the patient, she decid-
Fig. 1 (Michels and associates). Case 1. Preopera tive photograph of the left eye shows an epimacular membrane. Visual acuity was 20/125. ed to undergo pars plana vitrectomy and mem brane dissection, which was performed one week later. Intraoperatively, the epiretinal membrane was friable, adhered to the retina, and difficult to remove. The intraocular endoilluminator and light source were used. External illumination was provided by an operating mi croscope fitted with a 100-W halogen bulb transmitted with a fiberoptic cable. Total surgi cal time was one hour 50 minutes. Skin temper ature varied from 33.8 to 34.5 C. Oxygen satu ration was 99%. On the first postoperative day, her visual acuity was hand movements. By indirect ophthalmoscopy and biomicroscopy, a creamy oval area at the level of the deep retina and involving the fovea was noted. This area measured approximately 4 x 5 disk diameters. Photographs taken one week postoperatively demonstrated that this discrete lesion primarily involved the retinal pigment epithelium (Fig. 3). Fluorescein angiography performed two weeks postoperatively demonstrated discrete early mottled hyperfluorescence with late stain ing in the affected area (Fig. 4, left). Late stain ing of the overlying retinal vein could also be seen (Fig. 4, right). In March 1991, the visual acuity improved to 20/300 and the macular lesion showed mottled pigmentation. The pa tient was last examined in October 1991, at which time the visual acuity remained at 2 0 / 300.
Vol. 114, No. 3
Macular Phototoxicity
Fig. 2 (Michels and associates). Case 1. Preopera tive fluorescein angiogram shows normal retinal and choroidal vascular filling. Case 2 A 76-year-old w o m a n w h o h a d a five-month history of b l u r r e d vision in h e r r i g h t eye w a s e x a m i n e d in April 1 9 9 1 . H e r b e s t - c o r r e c t e d visual acuity w a s R.E.: 2 0 / 2 0 0 a n d L.E.: 2 0 / 2 0 . The anterior segment was normal. Examination of the r i g h t eye w i t h b i o m i c r o s c o p y a n d i n d i -
289
Fig. 3 (Michels and associates). Case 1. One week after vitrectomy, a creamy oval discrete lesion involv ing the outer retina and retinal pigment epithelium was seen in the macula region. Visual acuity was hand motions. rect o p h t h a l m o s c o p y s h o w e d a p r o m i n e n t m a c u l a r e p i r e t i n a l m e m b r a n e c a u s i n g fractional e l e v a t i o n of t h e fovea (Fig. 5). T h e r e w e r e p r o m i n e n t r e t i n a l striae a n d r e t i n a l m i c r o v a s cular a b n o r m a l i t i e s o n t h e s u p e r i o r e d g e of t h e
Fig. 4 (Michels and associates). Case 1. Postoperative fluorescein angiogram. Left, Early phase shows a discrete lesion with mottled fluorescence and hyperfluorescent margins. Right, Venous phase shows increased fluores cence and staining of the retinal vein overlying the lesion and at the level of the sclera.
290
AMERICAN JOURNAL OF OPHTHALMOLOGY
September, 1992
Fig. 5 (Michels and associates). Case 2. Prominent epiretinal membrane causing a tractional detachment of the fovea. Visual acuity was 20/200.
Fig. 6 (Michels and associates). Case 2. Three weeks postoperatively, the right eye demonstrated pigment clumping in the papillomacular bundle and surrounding the fovea. Visual acuity was 20/400.
epiretinal membrane. Pigmentary changes were not present. Fluorescein angiography also failed to show abnormalities at the level of the retinal pigment epithelium. On April 16, 1991, she underwent a three-port pars plana vitreetomy and membrane dissection in the right eye.
A thick and linear epiretinal membrane that extended across the macula was easily removed. Another epiretinal membrane on the inferonasal edge of the macula was also removed. The traction on the macula was released and no complications occurred. At the
Fig. 7 (Michels and associates). Case 2. Postoperative fluorescein angiogram. Left, Arteriovenous phase demonstrated mottled fluorescence of the lesion surrounding the fovea and in the papillomacular bundle. Right, Late phase of the angiogram shows staining of the area corresponding to the pigmentary disturbance noted clinically.
Vol. 114, No. 3
Macular Phototoxicity
end of the surgical procedure, cryotherapy was used to treat two preexisting anterior retinal tears, and a fluid-air exchange was performed. Total surgical time was one hour 50 minutes, and the intraocular endoilluminator and light source were used. External illumination was provided by a microscope fitted with a 50-W tungsten filament bulb. The patient's intraoper ative core temperature was 35.0 to 35.3 C. The oxygen saturation was 99%. One week postoperatively, visual acuity in the right eye was 20/400 and retinal examina tion disclosed macular edema. Three weeks postoperatively, hypopigmented mottling throughout the nasal portion of the macula and surrounding the fovea was noted at the level of the retinal pigment epithelium (Fig. 6). Fluores cein angiography showed early hyperfluorescence and late staining in the area where the pigmentary changes were seen clinically (Fig. 7). Four months postoperatively, the visual acu ity in her right eye had improved to 2 0 / 6 0 - 2 , although the macular pigmentary changes re mained unchanged. Case 3 A 72-year-old man with a visual acuity of 1/200 in the right eye underwent a pneumatic retinopexy for a macula-off rhegmatogenous retinal detachment on Aug. 26, 1988. On Sept. 29, 1988, visual acuity had improved to 20/80
Fig. 9 (Michels and associates). Case 3. Left, One ir demonstrated a persistent epimacular membrane. Vi shows persistent macular edema and no abnormalities
291
Fig. 8 (Michels and associates). Case 3. Four months after pneumatic retinopexy, an epimacular membrane was noted. Visual acuity was 20/200.
and the retina was totally attached. On Dec. 28, 1988, the visual acuity in the right eye de creased to 20/200 because of the development of an epiretinal membrane in the macular re gion (Fig. 8). On March 3, 1989, a pars plana vitrectomy and epiretinal membrane dissection
ith after vitrectomy for macular pucker, the right eye al acuity was 20/100. Right, Fluorescein angiogram the level of the retinal pigment epithelium or choroid.
292
AMERICAN JOURNAL OF OPHTHALMOLOGY
in the right eye were performed. On April 6, 1989, the visual acuity in the right eye was 20/100 and a persistent epiretinal membrane on the superotemporal macula that exerted traction on the fovea was noted (Fig. 9). On April 14, 1989, a repeat pars plana vitrectomy with epiretinal membrane dissection was per formed. The epiretinal membrane adhered to the retina and fragmented while it was being removed. No intraoperative complications oc curred. The Trek intraocular endoilluminator and light source were used. External illumina tion was provided by an operating microscope
September, 1992
fitted with a 100-W halogen bulb transmitted with a fiberoptic cable. Total surgical time was two hours 25 minutes. The oxygen saturation was 99 % andcore temperature was 35.5 to 36.4 C. Five days postoperatively, the visual acuity in the right eye was 20/70. The macula showed mild white opacification and edema of the outer retina involving the superior macula. On May 18, 1989, the visual acuity in the right eye was 20/80. Fundus examination disclosed a sharply demarcated area of hypopigmentation at the level of the pigment epithelium in the supero nasal macula (Fig. 10). Fluorescein angiogra-
Fig. 10 (Michels and associates). Case 3. Top left, One month after repeat vitrectomy for a persistent epimacular membrane, an area of hypopigmentation appeared superonasal to the fovea at the level of the retinal pigment epithelium. Visual acuity was 20/80. Top right, Early arteriovenous phase of the fluoresce in angiogram demonstrates minimal mottled hyperfluorescence. Bottom left, Venous phase of the fluo rescein angiogram demonstrates increased fluo rescence of the lesion superonasal to the fovea and mild macular edema.
Vol. 114, No. 3
Macular Phototoxicity
phy disclosed mottled hyper- and hypofluorescence corresponding to the area of retinal pigment epithelial hypopigmentation seen clin ically.
Results The epiretinal membranes in the macular region in our cases were secondary to a rhegmatogenous retinal detachment in one eye, a pos terior vitreous separation with the develop ment of retinal tears in one eye, and acute retinal necrosis in one eye. Preoperatively, the visual acuity ranged from 20/100 to 20/200. Intraoperatively, the epiretinal membranes ad hered to the retina and were difficult to remove in two of the three cases. In these two cases, the endoillumination was held close to the retina, and the surgical procedure was prolonged. Ret inal exposure to the light of the operating microscope, however, was limited to approxi mately 20 minutes for standard opening and closing associated with a three-port vitrectomy. In all three cases, macular lesions consisting of white opacification of the outer retina were noted within one week postoperatively. These findings were not noted intraoperatively. The macula was edematous, and in one eye (Case 1) acute segmental retinal phlebitis was also pres ent overlying the area of outer retinal opacifica tion. The macular lesions were all sharply de marcated, irregularly oval to round, varied in size from 2 to 5 disk diameters, and involved the fovea in two eyes. Initially, the visual acuity decreased in two eyes and improved in one eye. With time, visual acuity improved in the eyes with initial postoperative decrease in vision. The visual acuity of one eye improved to sub stantially better than the preoperative value. The time to recovery of best visual acuity varied from three to 12 months. With time, the white opacification of outer retina changed and was characterized by mottling of the retinal pig ment epithelium.
Discussion In our three cases, the characteristic course of the macular lesion and its fluorescein angio graphie features were typical of retinal photo toxicity.2 Phototoxic lesions are typically noted clinically only several days after the light expo
293
sure as in our three cases. A discrete margin of retinal pigment epithelial hypopigmentation is noted to develop as the initial outer retinal whitening fades. Fluorescein angiography clearly demonstrates the area of retinal pig ment epithelial disturbance, as early discrete hyperfluorescence throughout the area of mot tled pigmentary change evolves to late staining but never leakage. In most cases, the pigment clumping results in a monotonous, diffuse mot tled hyperfluorescence that can evolve into a target-like central clumping surrounded by hy popigmentation. Typically, lesions caused by the operating microscope assume the shape of the light source used. 6 The lesion produced by tungsten filaments often assumes the oval shape of the filament, whereas fiberoptic systems often gen erate a round lesion. 1 These lesions tend to have sharply demarcated borders partly because of minimal ocular or light position change during anterior segment procedures. The size of the lesion is usually less than 1 to 2 disk diameters. Our cases demonstrated the typical clinical and fluorescein angiographie features of opera ting microscope-induced retinal phototoxicity. The features that differentiate the lesions in our cases from the typical operating microscopeinduced lesion are the size and shape of the lesions and the minimal exposure to light from the operating microscope. In all three cases, the lesions were markedly larger than would be expected with the operating microscope light source. In Case 2, a nearly round lesion was seen despite the use of a tungsten filament in the microscope light source, which should re sult in a lesion of much smaller and more rectangular dimensions. The lesions in our cases had marked pleomorphism, probably be cause of intraoperative alterations in illumina tion positioning in terms of proximity to the retinal surface as well as location of illumina tion, which is typical of surface-level vitrec tomy, especially when it involves dissection of epiretinal membranes over a large area. The lesion in Case 3, however, had a fairly discrete epiretinal membrane above the fovea. Surface dissection involved a discrete area and the light exposure was likely limited largely to this area. This probably accounts for the comparatively discrete appearance of the phototoxic lesion. Similarly, anterior segment procedures which may result in operating microscope-induced retinal phototoxicity and experimental endoilluminator-induced retinal phototoxicity are characterized by minimal change in light
294
September, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
position and therefore may result in discrete lesions. 7 Retinal phototoxicity is a function of the following three factors: wavelength of the light source, duration of exposure, and power level. Duration of exposure and power level deter mine total energy delivered to a target tis sue. 8 For retinal phototoxicity to occur, energy must be transmitted to and absorbed by a target tissue. With an operating microscope, light is transmitted through the whole globe, a process that has been carefully studied. In the intact globe, reflection by and absorption in the tis sues anterior to the retina protects the retina from the short-wavelength ultraviolet light and some of the blue visible light that is most efficient in inducing phototoxicity. 910 With an endoilluminator, there is no absorp tion of damaging irradiation by anterior ocular structures once the vitreous is cleared of opaci ty. Fuller, Machemer, and Knighton 7 found that clinical phototoxicity can be seen after a 15minute exposure and histologie damage can be noted after just a 10-minute exposure in owl monkeys. In their carefully controlled study, monkeys were exposed to between five and 30 minutes of endoillumination, and then studied by ophthalmoscopy, fluorescein angiography, and light and electron microscopy at several intervals after exposure. As is true of all cases of photochemical damage, lesions were not immediately apparent even after the 30-minute exposure, but could be seen as early as five hours after exposure. This observation is criti cal for the vitreoretinal surgeon who is accus tomed to observing immediate thermal effects delivered by other forms of electromagnetic irradiation, such as the endolaser. 8 Kuhn, Morris, and Massey 4 suggested that the danger of iatrogenic photic damage from the endoilluminator probe with subsequent visual impairment is greatest when vitrectomy is per formed for epiretinal membrane dissection. They concluded that the problem is not widely recognized because the subtle changes clearly seen by fluorescein angiography are likely to be overlooked on ophthalmoscopy. We further suggest that many of the characteristic angio graphie findings noted at the level of the retinal pigment epithelium may not be present when retinal phototoxicity occurs in detached retina, as has been experimentally demonstrated by Zilis and Machemer. 11 In this instance, damage to noncontiguous retinal pigment epithelium may not be seen even with fluorescein angiog raphy.
Protection of the retina from endoilluminator-associated retinal phototoxicity is provided in part by macular xanthophyll. Jaffe and Wood12 were the most recent to demonstrate this qualitatively. Our Case 2 demonstrates macular sparing that may be explained in part by the macular xanthophyll. Ham and associ ates 8 demonstrated a twofold threshold energy increase required to produce a foveal compared to a parafoveal lesion in monkeys. Although no direct evidence links wave length attenuation with reduction of endoilluminator-associated retinal phototoxicity, ex perimental work in phakic dogs 13 and rhesus monkeys' 4 suggests that limitation of blue light exposure markedly reduces the incidence of indirect ophthalmoscope-induced and of oper ating microscope-induced retinal phototoxi city, respectively. Although Fuller, Machemer, and Knighton 7 documented less than 5% blue light output from their light pipe endoilluminator, and Meyers and Bonner' 5 measured that the blue fraction is less than 3 % of transmitted energy, both groups of authors emphasized that elimination of light less than 500 nm from fiberoptic endoilluminators may decrease the risk of retinal phototoxicity by tenfold. Experimental proof of the benefit of blue light filtration in the endoilluminator probe is lacking. Other factors that are somewhat controllable are the power level of the light source and the exposure time. The currently popular cooler (and bluer) light sources, such as the xenon ~ 20 CM
E E w 15 CO
5 = 10
î
0
«-
10
15
20
25
Distance From Endoilluminator tip (mm)
Fig. 11 (Michels and associates). Endoilluminator irradiance as a function of the distance from the light source (mW/mm2) in air measured using the Trek endoilluminator and light source.
Vol. 114, No. 3
Macular Phototoxicity
source, and the higher intensity often needed for video recording equipment need to be con sidered when performing a surgical procedure. Maximum output of the Trek endoilluminator and light source and the Grieshaber endoillu minator and light source was determined to be 15.6 and 18.0 mW/mm 2 , respectively, using a silicone diode photodetector (UDT model 247) traceable to the National Bureau of Standards. Although power level is not often considered a flexible variable, it is noteworthy that the reti nal irradiance value is inversely proportional to the square of the distance of the source light to the retinal surface.16 Thus, moving the light source as far away from the retinal surface as is possible during noncritical periods of mem brane dissection seems prudent (Fig. 11). With regard to exposure time, some investigators have advocated intermittent exposure as less likely to cause the same degree of damage as constant exposure for a given cumulative time period. 4 The importance of intermittent expo sure is debatable. Data on brief intermittent exposures in pseudophakic primates 17 and phakic primates 18 contend that these exposures are cumulative and that intermittent intraoper ative rest periods are unlikely to be protective. There are two variables independent of the immediate ocular milieu that probably do have a role in retinal phototoxicity. System tempera ture has long been recognized as important in retinal phototoxicity. The idea that the photo chemical process might be enhanced by in creased core temperature was addressed and shown by Noell and associates. 19 Rinkoff and associates 20 demonstrated that cooling infusion fluid from 39 C to 29 C in rabbits in which vitrectomy had been performed while closely illuminating the retina with an endoilluminator increased the damage threshold time from 25 to 60 minutes. Although neither alteration of core temperature nor use of hypothermie infusion fluid has been studied in humans, it is prudent to avoid unnecessary increases in core or infu sion fluid temperature. The other variable not immediately associat ed with the globe is tissue oxygénation. Be cause the photochemical process is thought to result from the action of free radical and singlet oxygen moieties, high levels of inspired oxygen might decrease the threshold for phototoxicity. Jaffe and associates 21 studied the role of in spired oxygen in phakic rhesus monkeys ex posed to operating microscope light under con ditions of 2 1 % Fi0 2 and 99% Fi0 2 and noted marked potentiation of phototoxic damage un
295
der conditions of increased oxygénation. Al though oxygénation has not been directly stud ied under conditions of endoilluminatorrelated retinal phototoxicity, it seems prudent to avoid unnecessary oxygénation in patients undergoing a surgical procedure using the en doilluminator. Endoilluminator-related retinal phototoxic ity is a largely unrecognized, and partially preventable, sequela to certain cases of vitrec tomy. Awareness of the potential for its occur rence and minimization of both intraocular and extraocular variables that may contribute to the process are advised.
References 1. McDonald, H. R., and Harris, M. J.: Operating microscope induced retinal phototoxicity during pars plana vitrectomy. Arch. Ophthalmol. 106:521, 1988. 2. Michels, M., and Sternberg, P.: Operating mi croscope-induced retinal phototoxicity. Pathophysiology, clinical manifestations and prevention. Surv. Ophthalmol. 34:237, 1990. 3. McDonald, H. R., Verre, W. P., and Aaberg, T. M.: Surgical management of idiopathic epiretinal membranes. Ophthalmology 93:978, 1986. 4. Kuhn, F., Morris, R., and Massey, M.: Photic retinal injury from endoillumination during vitrec tomy. Am. J. Ophthalmol. 111:42, 1991. 5. Kelly, N. E., and Wendel, R. T.: Vitreous sur gery for idiopathic macular holes. Results of a pilot study. Arch. Ophthalmol. 109:654, 1991. 6. Khwarg, S. G., Linstone, F. A., Daniels, S. A., Isenberg, S. J., Hanscom, T. A., Geoghegan, M., and Straatsma, B. R.: Incidence, risk factors, and mor phology in operating microscope light retinopathy. Am. J. Ophthalmol. 103:255, 1987. 7. Fuller, D., Machemer, R„ and Knighton, R. W.: Retinal damage produced by intraocular fiber optic light. Am. J. Ophthalmol. 85:519, 1978. 8. Ham, W. T„ Ruffolo, J. J., Mueller, H. A., and Guerry, D.: The quantitative dimensions of intense light damage as obtained from animal studies. The nature of retinal radiation damage. Dependence on wavelength, power level and exposure time. Vision Res. 20:1105, 1980. 9. Boettner, E. A., and Wolter, J. R.: Transmission of the ocular media. Invest. Ophthalmol. Vis. Sci. 1:776, 1962. 10. Ham, W. T., Mueller, H. A., Ruffolo, J. J., and Clarke, A. M.: Sensitivity of the retina to radiation damage as a function of wavelength. Photochem. Photobiol. 29:735, 1979. 11. Zilis, J. D., and Machemer, R.: Light damage in detached retina. Am. J. Ophthalmol. 111:47, 1991. 12. Jaffe, G. J., and Wood, I.: Retinal phototoxicity
296
AMERICAN JOURNAL OF OPHTHALMOLOGY
from the operating microscope. Protective effect of the fovea. Ophthalmology 106:445, 1988. 13. Buyukmihci, N.: Photic retinopathy in the dog. Exp. Eye Res. 33:95, 1981. 14. Weil, L., Michels, M., Sternberg, P., and Aaberg, T. M.: Filtration of blue (<515nm) light significantly reduces retinal phototoxicity. ARVO ab stracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1989, p. 459. 15. Meyers, S., and Bonner, R. F.: Retinal irradiance from vitrectomy endoilluminators. Am. J. Oph thalmol. 94:26, 1982. 16. Calkins, J. L., Hochheimer, B. F., and D'Anna, A.: Potential hazards from specific ophthalmic devic es. Vision Res. 20:1039, 1980. 17. Irvine, A. R., Wood, I., and Morris, B. W.: Reti nal damage from the illumination of the operating
September, 1992
microscope. An experimental study in pseudophakic monkeys. Arch. Ophthalmol. 102:1358, 1984. 18. Lawwill, T., Crockett, R. S., and Currier, G.: Retinal damage secondary to chronic light exposure. Thresholds and mechanisms. Doc. Ophthalmol. 44:379, 1977. 19. Noell, W. K., Walker, V. S., Kang, B. S., and Berman, S.: Retinal damage by light in rats. Invest. Ophthalmol. 5:450, 1966. 20. Rinkoff, J., Machemer, R., Hida, T., and Chan dler, D.: Temperature-dependent light damage to the retina. Am. J. Ophthalmol. 102:452, 1986. 21. Jaffe, G. J., Irvine, A. R., Wood, I. S., Severinghaus, J. W., Pino, G. R., and Haugen, C : Retinal phototoxicity from the operating microscope. The role of inspired oxygen. Ophthalmology 95:1130, 1988.