Case reports to accompany Early Treatment Diabetic Retinopathy Study Reports 3 and 4. The Early Treatment Diabetic Retinopathy Study Research Group.

The principles exemplified by the preceding case reports are summarized below. 1. Although the presence of DRS high-risk characteristics is the single most important indication for initiating scatter photocoagulation, intraretinal lesions suggesting ischemia (soft exudates, IRMA, venous beading, arteriolar abnormalities, and moderately severe hemorrhages and/or microaneurysms) are also important. When these lesions are severe, rapid progression is likely, and initiation of scatter photocoagulation should be considered for at least 1 eye, even when new vessels are absent or mild (Cases 3, 5, 9, and 11). Both eyes should be followed carefully, whether treated or not, and special attention to blood pressure and renal status may be important. When these intraretinal lesions are mostly absent or mild, progression of PDR may be very slow (Cases 1, 6, and 7). 2. NVD are the single most important prognostic feature of diabetic retinopathy, and when they are well established (i.e., greater than or equal to DRS Standard Photograph 10A), the indication for initiation of scatter photocoagulation is strong (Cases 7, 8, 10, and 11). 3. NVE in the absence of vitreous or preretinal hemorrhage or the severe intraretinal lesions listed in item 1 are a weaker indication for photocoagulation, and careful observation of such eyes is a reasonable alternative to prompt treatment. 4. The initial vitreous or preretinal hemorrhage in eyes with PDR is rarely so large that photocoagulation cannot be carried out before a subsequent larger hemorrhage occurs, provided patients report symptoms and are examined promptly. In such cases it is prudent to treat the lower quadrants first, if possible, before they become obscured by hemorrhage (Cases 3-7 and 9-11). 5. Even after full scatter photocoagulation, with burns placed no more than one-half burn diameter apart, there is ample room for additional treatment between scars, and this often seems to be effective in encouraging regression of new vessels that remain or recur after the completion of the initial treatment. Such additional scatter treatment may be concentrated in areas of NVE (Cases 2 and 5) or applied throughout the fundus (Cases 4 and 9-11). Extension of scatter photocoagulation into the posterior pole also appears to be effective sometimes (Case 8). 6. Knowledge of the tendency for new vessels to follow a cycle of proliferation and regression is important when considering additional scatter treatment when new vessels fail to regress or recur after initial scatter treatment.(ABSTRACT TRUNCATED AT 400 WORDS)