Prophylaxis in Fellow Eye of Primary Retinal Detachment What Not to Do and What to Do

Norman Byer

It is generally very helpful in understanding the present to make a retrospective survey of the thinking of the past, which has led us to our present concepts. The progression of ideas in the case of prophylactic treatment of retinal detachment first developed from a few correct elementary clinical observations, but then proceeded on the basis of mostly theoretical reasoning because of the profound dearth of empirical data.

Along with the early realization that some retinal detachments could be successfully treated, other observations also began to be made. Certain associated pre-existing retinal lesions began to be observed in eyes in which causative retinal tears had led to retinal detachment. It was thought that perhaps these associated lesions were responsible for the onset of the retinal tears and, therefore, were of prognostic importance. With the advance of more careful retinal examination, more of these lesions began to be discovered in the "fellow eyes" of patients who had had a retinal detachment in their primary eyes and also in the eyes of patients who had not suffered a retinal detachment.

The idea soon began to be entertained that perhaps retinal detachments could be prevented, in the first instance, by "treating" these associated lesions before the initiating retinal tears occurred. Because natural history information regarding these various retinal lesions was almost non-existent at that time, the concept of the value of such "prophylactic" treatment rested on purely theoretical reasoning. This gulf between theoretical expectations and actual empirical data was easily, and unknowingly, bridged by several broad assumptions which were as follows: (1) the occurrence of bilateral retinal detachment was thought to be in the range of 20-50% of patients who had suffered a primary detachment, (2) the associated pre-existing "suspect" retinal lesions were thought to represent the precursor sites from which retinal tears would later arise, and (3) the pretreatment of these pre-existing visible retinal lesions was thought to prevent later retinal tears and detachment.

This thinking, which was believed to justify the concept of "prophylactic" treatment,was greatly advanced and crystallized in the 1950s by the bringing together of two important developments in ophthalmology. The first of these was the popularization by Schepens [1] of a definitive method of retinal examination, using binocular indirect ophthalmoscopy combined with simultaneous localized scleral indentation. This method opened the possibility of examining in detail all areas of the retina in multiple, stereo-scopically viewed images. This development was a vast advance over previous methods and eventually led to the accurate characterization of various peripheral retinal lesions and to the quantitative collection of natural history data that had previously remained unknown.

The second very significant event that influenced the thinking regarding "prophylactic" treatment was the invention by Meyer-Schwickerath [2] of an effective method to deliver controlled photocoagulation energy to the retina to produce discrete retinal burns which later became converted into small scars.

The concurrence of these two events had the momentous effect of opening up the vastly improved possibilities of finding and of treating many peripheral retinal lesions which heretofore had remained hidden. In a relatively short time, as the result of very successful promotion and distribution, photocoagulation instruments became available throughout the world. It is very easy to understand, based on the uncritical acceptance of the previously mentioned three assumptions, that a new "standard of care" soon emerged throughout the world that prevention of retinal detach ment could (and should) be achieved by systematic "prophylactic" treatment of various pre-existing asymptomatic retinal lesions.

Two large long-term surveys of reports in the literature, purporting to substantiate the correctness of this view were published by Meyer-Schwickerath and Fried in 1980 [3] and by Haut et al. in 1988 [4] all of whom were staunch advocates of this standard of care and believed that it provided substantial success in achieving the goal of preventing retinal detachment. Both surveys revealed that there was a residual risk of retinal detachment, even after those attempts to prevent it, amounting to 5% in the first report [3] and 2-5.5% in the second report [4], depending on the modality used.

Eventually, however, various reports began to appear which tended to agree in showing that the three underlying assumptions which formed the basis of the new standard of care were not accurate. With regard to the first assumption, the bilaterality of retinal detachment had been considerably overestimated, and instead of being 20-50%, was in the range of 6-11% [5-13] (Table 2.1).

With regard to the second assumption, it has been reported that 72% of new symptomatic retinal tears occur in retinal areas that appear clinically normal [14]; and, in a large autopsy study of eyes

Table 2.1. Incidence of bilateral retinal detachment


Incidence (%)

Toernquist 1963 [5]


Edmund 1964 [6]


Boeke 1966 [7]


Michaelson et al. 1969 [8]


Davis et al. 1974 [9]


Bleckman and Engels 1975 [10]


Haut and Massin 1975 [11]


Laatikainen and Harju 1985 [12]


Toernquist et al. 1987 [13]


Table 2.2. Remaining risk of retinal detachment (RD) following "prophylactic" treatment of fellow eyes with predisposing lesions


Risk of RD (%)

Michaelson et al. 1972 [16]


Morax et al. 1974 [17]


Dralands et al. 1980 [18]


Meyer-Schwickerath and Fried 1980 [3]


Girard et al. 1982,1983 [19,20]


Haut et al. 1988 [4]


Folk et al. 1989 [21]


Table 2.3. Incidence of retinal detachment in fellow eyes of comparison groups of patients with "dangerous" lesions without and with "prophylactic" treatment


Without Rx (%)

With Rx (%)

Dralands et al. 1980 [18]



Girard et al. 1982,1983 [19,20]



Folk et al. 1989 [21]



with lattice degeneration, 79% of the tears were located in such areas [15].

As for the third assumption, various reports have shown the still remaining rate of detachment following "prophylactic" treatment of fellow eyes to be from 2% to 9% [3,4,16-21] (Table 2.2).

It is especially helpful in this discussion to present data reported by authors who compared two parallel groups of patients - one being treated and one not being treated [18-21]. These are summarized in Table 2.3.

This led Michaelson et al. [16] to say that "no notable drop in fellow eye detachment had occurred", and they officially discontinued the practice of "prophylactic" treatment. Dralands et al. [18]

also concluded that "the incidence of second eye detachments does not decrease as the result of preventive treatment". These data are summarized in Table 2.2.

The special category of aphakic or pseudophakic fellow eyes has very little pertinent data from comparative studies in the literature comparing treated and untreated groups. However, such a study was reported in 1989 by Herzeel et al. [22] in which one of the groups was treated with encircling circumferential cryotherapy. They found that the treated group developed retinal detachment in 2.3%; whereas, this outcome occurred in only 1.3% of the untreated group, leading the authors to say that "these results lead us to conclude that 'prophylactic' treatment does not necessarily prevent this complication".

A further category of fellow eyes that simultaneously harbor multiple risks has always represented a special group which has been thought to have a more marked vulnerability to detachment and therefore to be pre-eminently eligible for "prophylactic" treatment. However, Folk et al. [21] in 1989 reported comparison groups of fellow eyes, all of which had three simultaneous risk factors for retinal detachment (fellow eye status, lattice degeneration and high myopia). They found that "prophylactic" treatment of the lattice lesions did not confer any advantage in lowering the rate of detachments. Instead, apparently what happens is that, although the risk of detachment is higher with existing multiple risk factors, so also is the risk of secondary detachment following treatment.

In fact the incidence of retinal detachment following "prophylactic" treatment is in approximately the same range as the rate of detachment in fellow eyes left untreated.

Therefore, we may conclude that the earlier hope of preventing retinal detachment in fellow eyes by some form of "prophylactic" treatment has not been significantly substantiated, and this approach offers no more than a slight benefit.

The significant visible predisposing lesions of the peripheral retina, related to retinal detachment, which are primarily lattice degeneration, senile retinoschisis, retinal breaks, and cystic retinal tufts also have a significant prevalence rate in primary eyes. Here, also the natural histories as well as the futility of applying so-called "prophylactic" treatment has been amply documented [23-26].

In summary, we may say that the well-established practice of applying "prophylactic" treatment to visible predisposing peripheral retinal lesions, whether in primary eyes or in fellow eyes, before any detachment has occurred, and specifically for the purpose of preventing this outcome has by now been thoroughly discredited and must be discarded as the "standard of care". This answers the first major question of this chapter, and constitutes, "What NOT to do".

This does not mean however that retinal detachments cannot be prevented. However, our clinical attention must be directed to a completely different matter. We should not treat asymptomatic eyes (whether primary or fellow eyes), but we should be on the lookout for patients who complain of recent visual symptoms that suggest the occurrence of a posterior vitreous detachment. Such symptoms consist of the sudden appearance of vitreous floaters or light flashes. It has been reported that among patients older than 50 years of age, the symptom of suddenly appearing floaters is known to be caused by posterior vitreous detachment in 95% of cases [27].

Patients with this complaint all should be thoroughly and conscientiously examined with indirect ophthalmoscopy and simultaneous scleral indentation in both eyes, to search for any new trac-tional retinal tear or tears that may be present. If such are discovered they should be promptly treated by surrounding them with either laser photocoagulation or cryotherapy. It has been found in several clinical series [14,28-31] and in an autopsy series by Foos [32] that about 15% of eyes that have had a vitreous detachment also have a tractional retinal tear or tears. It has been reported that around 28% of these will progress to a retinal detachment before the patient first consults an ophthalmologist [33].

If the remaining 72% of eyes with fresh retinal tears are not promptly discovered and treated, about one-third of this number will progress to a retinal detachment either very soon or within 2-3 months. Among eyes discovered with fresh tractional retinal tears but which have not yet led to retinal detachment, it has been well established that about 35% will lead to retinal detachment [34-36].

Therefore, prompt examination of patients with suggestive symptoms of vitreous detachment, followed by prompt treatment of any new tractional retinal tears, is mandatory and will provide a high probability of preventing further progress to a retinal detachment, which is a much more serious event. Altogether these account for approximately 95% of all retinal detachments. If all fresh retinal tears resulting from posterior vitreous detachment were discovered at the time when no retinal detachment was yet present, and were successfully treated at that time, then approximately 44% of retinal detachment in general could be prevented (Table 2.4). We must remember that about two-thirds of the eyes with such tears would never progress to detachment even without treatment. Herein lies a very great opportunity for the successful prevention of this historic and continuing scourge of vision.

It may be thought that eyes that undergo a sudden posterior vitreous detachment may have a still higher risk if they contain pre-existing visible predisposing lesions related to retinal detachment. However, in a large prospective study of such eyes [14] it was

Table 2.4. Fate of new retinal tears resulting from posterior vitreous detachment

Progressing promptly to retinal detachment


(before consulting ophthalmologist)

Expected to progress to detachment


(without treatment) = 72/3

Total progressing to retinal detachment


(without treatment)

Proportion of retinal detachments prevented (in general)


found that eyes with pre-existing retinal breaks do not have any increased risk of retinal detachment following PVD. The same was true of eyes with pre-existing senile retinoschisis (N.E. Byer, unpublished observations). In the group of eyes with pre-existing lattice degeneration the onset of sudden PVD did lead to new retinal tears in 24% (N.E. Byer, unpublished observations). In 76% of the eyes with lattice degeneration, the occurrence of a PVD did not produce any complication of any kind (N.E. Byer, unpublished observations). However, in 50% of the lattice-eyes the new tears appeared in "normal"-appearing retinal areas and, thus, could not have been pretreated (using the "prophylactic" method) because they would not have been detected.

In summary, the conscientious examination of patients with recent symptoms of posterior vitreous detachment, followed by discovery and treatment of new retinal tears, provides the capability of preventing around 44% of rhegmatogenous retinal detachments. The older practice of "prophylactically" treating eyes with predisposing retinal lesions has led only to a disappointing and questionable clinical benefit. That procedure should be officially discarded, and the above much-preferred method should be recognized as the proper standard of care. This then is the obvious and available answer to the question of "What to DO" to prevent retinal detachment.

0 0

Post a comment