without subsequent detachment of the retina.

Visual acuity was converted to decimal notation and plotted as worse, unchanged, or improved (Fig 12). The visual acuities of eyes in the control group appear clumped predominantly in the "worse" category. Statistical analysis by the chi-square test, however, reveals no significant difference from the eyes treated by coagulation (Table 6A). However, if the visual acuities are divided as to those which either improved or did not decrease more than two Snellen lines, vs those which did decrease

more than two Snellen lines, the eyes in the control group fared significantly worse than did the eyes treated by coagulation (Table 6B).

The peripheral visual field isopters were somewhat constricted by photocoagulation in 12 of the 27 eyes (44%). In no constriction noticeable to the patient, such instance, however, was the field except under circumstances of perimetric testing. Because most treatments of neovascular lesions occurred in the anterior equatorial regions of the retina, often superotemporally,1,2 corresponding field defects remained peripheral and often occurred inferonasally, an

551

In a series of eyes observed previously by the author, and followed without intervention for an average of 31 months, the percent of eyes involved by stage III lesions increased slightly from 68% to 72%, and the percent of eyes involved by stage IV lesions (vitreous hemorrhage) increased from 28% to 44%. Many of the eyes included in that study of the natural history of PSR are reported again here, and they serve as the control group. Furthermore, it is of importance to the interpretation of the results of photocoagulation that no eye has shown spontaneous regression

of neovascular lesions. The natural course of PSR thus appears to reflect a progressive, sequential, and pathogenetically related chain of events, leading ultimately, in many cases, to retinal detachment (stage V). Consequently, therapeutic and prophylactic intervention appear warranted prior to the onset of those factors, such as neovascularization and vitreous hemorrhage, that apparent

0

P < 0.025

ly lead to retinal detachment. Following intervention with photocoagulation, the reduction of the number of eyes in this series with neovascular lesions (from 100% to 55%) and the reduction of eyes with vitreous hemorrhage (from 41% to 26%) appear to represent a favorable reversal of the natural course of events in PSR; without intervention, stages III and IV, as shown by the control group, characteristically either stayed the same or worsened, but did not improve spontaneously. It also must be stressed that many eyes reported here, which have residual patent sea fans, are still undergoing photocoagulation therapy, and with additional follow-up, are likely to show complete absence of residual stage III lesions.

There appears to be little, if any, difference between the results of therapeutic photocoagulation and those of prophylactic photocoagulation. However, one should not necessarily conclude from

554

Visual acuity prior to treatment appeared largely unaffected by PSR, including the visual acuity of those eyes with spontaneously occurring hemorrhages into the vitreous. The explanation for the retention of visual acuity in the presence of advanced retinopathy lies in the characteristic retinal geography of PSR. Neovascular lesions tend to arise preferentially in the equatorial region of the temporal quadrants of the fundus. Initial vitreous hemorrhages usually do not break into the center of the vitreous cavity, and thus the symptomatology may include "spots," "flashes," "floaters," and so on, but not usually a significant reduction in central acuity. Moreover, the equatorial lesions are peripheral enough that constriction of the visual field is not apparent to the patient and cannot usually be demonstrated in perimetric visual determinations.

Complications incurred through the use of photocoagulation did not seem excessive, and visual acuity was not adversely affected. The most common complication was peripheral field constriction of a symptomatically insignificant type. Maculopathy did not occur, possibly because of the peripheral location of the photocoagulation.' However, one case of maculopathy did occur in the control group, probably because of a traction band leading from a bleeding neovascular lesion to the macula.2 Vitreous hemorrhage, which could be at

TR AM ACAD OPHTH & OTOL

tributed to the effects of photocoagulation, occurred three times in a total of 58 treatment episodes. This rate of vitreous hemorrhage seems clinically acceptable, when one considers that the untreated natural course of events includes a definite rate of spontaneous vitreous hemorrhage (from a value of 28% initially to 44% two years later in the series of untreated eyes),2 and when one considers that photocoagulation in the eyes treated was able to eliminate the neovascular sources of vitreous hemorrhages 74% of the time (Table 1).

Because of the apparently favorable results reported herein, and because of the relative ease with which small, flat neovascular lesions are obliterated, it would seem advantageous to photocoagulate stage III lesions as soon as they are detected, even if vitreous hemorrhage has not yet occurred. If one permits a sea fan to remain perfused with blood, definite risks of progressive neovascularization and vitreous hemorrhage are incurred.2 Furthermore, larger sea fans, when photocoagulated, are more likely to bleed and are more difficult to occlude than are smaller ones.

As new sea fans make their appearance, the present evidence suggests that repeat photocoagulation may well be in the patient's best interests. In one respect, PSR resembles proliferative diabetic retinopathy." In both types of retinal disorders, the presence of a systemic disease is responsible for the continued development of retinal neovascularization.3 Consequently, such retinas cannot necessarily be considered permanently cured, even if all visible neovascular lesions can be eliminated at any one time. Fluorescein angiography should be performed at regular intervals on such patients. Persistent perfusion of neovascular tissue with fluorescein is evidence of incomplete treatment.

VOLUME 75
MAY-JUNE 1971

SUMMARY

Twenty-seven eyes with PSR were treated, primarily by xenon arc photocoagulation. The average follow-up duration was 30.4 months. The number of retinas with neovascularization was reduced from 100% to 55%, and the number of eyes with vitreous hemorrhage was reduced from 41% to 26%. Of the total number of neovascular lesions, 74% were obliterated by photocoagulation. The reduction in extent of retinal neovascularization was statistically significant when compared with an untreated control group of eyes. Intervention thus appeared to interrupt the natural course of PSR.

Fluorescein angiography is essential in the evaluation of completeness of photocoagulation, because perfused neovascular channels may be invisible with ophthalmoscopy alone.

Elimination of neovascular lesions by photocoagulation appeared safer and more likely to occur when these lesions were small and lay flat against the retina.

Vitreous hemorrhage following photocoagulation appeared more likely to occur when individual neovascular lesions were enlarged and elevated.

Photocoagulation did not adversely affect vision and, within limits, may have preserved it. Peripheral field constriction results from photocoagulation, but it is symptomatically insignificant.

ACKNOWLEDGMENT

Photographic assistance was provided by Messrs. Terry George and Norbert Jednock. Statistical assistance was provided by Meth Linwong, MD. Diagnosis of sickling disorders was confirmed by the Special Hematology Laboratory, Department of Medicine, Johns Hopkins University School of Medicine, sup

555

4. Hannon JF: Vitreous hemorrhages: Associated with sickle cell-hemoglobin C disease. Amer J Ophthal 42:707-712 (Nov) 1956

5. Newcomb NS, Goldberg MF, Welch RB: Intravenous fluorescein photography of the ocular fundus in sickle-cell anemia. Med Biol Illus 17:95-98 (April) 1967

6. Okun E: Development of sickle cell retinopathy. Docum Ophthal 26:574-581, 1969

7. Okun E, Cibis PA: The role of photocoagulation in the therapy of proliferative diabetic retinopathy. Arch Ophthal 75:337-352 (March) 1966

8. Romayananda N, Green WR, Goldberg MF: Histopathologic study of sickle cell retinopathy including trypsin digestion study. Presented in part at the meeting of the Association for Research in Ophthalmology, Sarasota, Fla, May 5, 1970 (To be published)

9. Ryan SJ, Goldberg MF: Anterior segment ischemia following scleral buckling in sickle cell hemoglobinopathy. Amer J Ophthal (In press)