Changing Concepts
of Failures
After Retinal Detachment William F.
•
Rachal, MD, Thomas C. Burton,
One thousand eighty-eight
consecu¬
tive
operations for retinal detachment were analyzed to determine the influence of current methods of examination and treatment on failures following surgery. The majority of failures were produced by preretinal membrane formation (33%) and massive preretinal retraction (27%). Other causes of failure included undetected reti¬ nal breaks (13%), inadequate scierai buckle (10%), new retinal breaks (8%), inadequate chorioretinal reaction (7%), and iatrogenic retinal breaks (2%). Pri¬ mary operations yielded an initial cure rate of 76%. Successful reoperations raised the final cure rate to 89%. Reoper¬ ations were associated more frequently with preretinal membrane formation and had a lower final cure rate (83%). This study emphasizes the importance of recognition and management of preretinal membrane formation. Except for new reti¬ nal tears and massive preretinal retrac¬ tion, surgical failures can be avoided by improved utilization of current examina¬ tion and operative techniques.
(Arch Ophthalmol 97:480-483, 1979)
few decades ago failures after "^ retinal detachment surgery often were attributed to high myopia, apha¬ kia, extent of the detachment, and inflammation.1 Inability to localize retinal breaks was not regarded as noteworthy. Gradually there has de¬ veloped the realization that an inade¬ quate surgical technique results in the Accepted
for publication June 8, 1978. From the Department of Ophthalmology, University of Iowa Hospitals and Clinics, Iowa
City. Reprint requests to C. S. O'Brien Library, Department of Ophthalmology, University of Iowa Hospitals and Clinics, Iowa City, IA 52242 (Dr Burton).
Surgery
MD
majority of failures in retinal surgery. In 1964 Okamura et al2 emphasized that the fundamental problem in
unsuccessful retinal surgery is the inability to close all the retinal tears contributing to the detachment. In addition, there has been an increasing respect for proliferative tissue on or beneath the retina that frequently prevents closure of retinal tears. With the popularization of wider scierai buckling materials and the development of improved examina¬ tion techniques with the binocular indirect ophthalmoscope, scierai in¬ dentation, and the peripheral fundus
lens, surgical success rates improved steadily. This study
contact
have
undertaken to determine the influence of current methods of exam¬ ination and treatment on the causes of failure following retinal detachment was
surgery. SUBJECTS AND METHODS One thousand eighty-eight consecutive cases of rhegmatogenous retinal detach¬ ment operated on at the University of Iowa
Hospitals and Clinics between January April 1973 were evaluated. Not included in the study were 205 patients with retinal breaks without a surrounding 1969 and
detachment. A retinal detachment was defined as at least 1 disc diameter of fluid surrounding the margins of a retinal break. Surgical failure was defined as the continued presence of subretinal fluid that was not confined to the peripheral fundus or was not stable. Surgical techniques consisted of lamellar scierai dissection with buried solid silicone rubber implants (61%), solid silicone rubber expiants (12%), silicone sponge rubber expiants (7%), demarcation with photoco¬ agulation or cryopexy (9%), combinations of the preceding operations (10%), and
other procedures (1%). Cryopexy was used in all but 11 cases of scierai buckling oper¬ ations. Eighty-two percent of the opera¬ tions used an encircling silicone band. Data on drainage of subretinal fluid were not available. There were 863 primary cases; 190 cases required a total of 225 additional operative procedures. One hundred fifty-six patients underwent two or more scierai buckling procedures. Twelve patients had a scierai buckle followed by argon laser or xenon photocoagulation to contain residual sub¬ retinal fluid. Eighteen patients had a primary demarcation procedure by cryocoagulation or photocoagulation followed by a scierai buckling operation. Four addi¬ tional patients were treated by multiple attempts at demarcation without scierai
buckling. Photocoagulation
to increase the chorio¬ retinal reaction on the buckle during the immediate postoperative period was con¬ sidered part of the primary operation. Sixty-four cases treated in this manner were not considered as reoperations.
RESULTS
The primary cause for each of the 259 failures following the initial reti¬ nal detachment procedures is tabu¬ lated as follows. The causes of failure of subsequent operations are not included since preretinal membrane formation becomes disproportionately more common with multiple reopera¬ tions. Cause of Failure
No. of Cases
(%)
Massive preretinal retraction 69(26.6) Preretinal membrane 85(32.8) Undetected retinal tear 34(l3.1) Inadequate scierai buckle 25(9.7) New retinal tear 20(7.7) Inadequate chorioretinal reaction 18(6.9) Iatrogenic retinal tear 5(1.9) Loss of buckle height 2(0.8) Macular hole 1(0.4) Total
Downloaded From: http://archopht.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/26/2015
259(99.9)
Inadequate closure of known retinal breaks was by far the most common cause of failure of the primary proce¬ dures, accounting for 77% of all primary failures and 69% of reoperations. Massive Preretinal Retraction
Massive preretinal retraction (MPR) is regarded as synonymous for the older term "massive vitreous retraction" (MVR) and the newer term "massive periretinal prolifer¬ ation" (MPP).-'4-"' In this situation extensive proliferative tissue, proba¬ bly derived from retinal glial and pigment epithelial cells, results in a "morning glory" detachment, which is inoperable by current methods of scierai buckling, or reduces visual function to light perception.57 Massive preretinal retraction devel¬ oped in 69 of 863 primary cases (8%), which were regarded subsequently as inoperable. The risk of MPR rises to after 13.8% two operations (P
.05).
Preretinal Membrane Formation
Preretinal
membrane
formation
histologically is probably identical to the proliferative tissue that produces inoperable MPR. Clinically the mem¬ brane is expressed in less severe form and is recognized by the appearance of rolled-edge tears, star folds, equa¬ torial ridges, and fixed radial folds. In this series preretinal membrane for¬ mation was observed preoperatively in 68% of patients undergoing reoperations, which was significantly higher than the 28% rate observed in primary cases (P < .01). Surgical failure was not ascribed to preretinal membrane formation un¬ less the choice and position of implants were technically satisfactory and no other contributing factors =
could be determined. Fishmouth tears, stiff radial folds communicating with posterior fluid, and tears that reopen¬ ed on the buckle without loss of buckle height were included in this catego¬ ryUndetected Retinal Breaks
Retinal breaks that are not ob¬ served preoperatively may be grouped in three categories: (1) missed retinal breaks that were present preopera¬ tively; (2) iatrogenic breaks produced at surgery; and (3) new retinal tears
arising postoperatively. Iatrogenic breaks occur most com¬ monly during drainage of subretinal fluid, occasionally at sites of anchor¬ ing sutures for encircling elements, or
while performing lamellar dissection in thin sclera. Presentation of formed vitreous almost invariably indicates the production of a new retinal break. An exception to this may occur if a drainage site is selected beneath a large retinal break or a giant tear through which vitreous can enter the sclerotomy. Usually the surgeon is aware of such a complication and can perform immediate remedial proce¬ dures to prevent subsequent reoperations. Conversely, the retina can be perforated by a deep penetration of a needle or diathermic tip without vitreous loss. This complication, espe¬ cially if posterior to the scierai buckle, is the usual cause of reoperations from iatrogenic breaks. A retinal break was considered new if the area in which it occurred were adequately examined preoperatively, if it were completely attached in the early postoperative phase, or if the tears were large or multiple and unlikely to have been missed preoper¬
atively.
All other tears that were not observed before the first operation considered missed retinal were breaks. Failures were attributed to missed retinal breaks regardless of any other contributing factor, includ¬ ing preretinal membrane formation.
Inadequate Scierai Buckle The position or size of the scierai buckle was implicated in any case in which the entire retinal break was not isolated on the apex or plateau of the buckle and fluid around the break communicated with the residual reti¬ nal detachment. In addition, the choice of implant was implicated if it were believed to be of insufficient height or width to compensate for the amount of preretinal membrane for¬ mation known to exist preoperative¬
ly.8·9
Loss of Buckle
Height
Loss of buckle height may produce new tears or reopen previously sealed tears. A typical example of this prob¬ lem is recurrent detachment following removal of a scierai buckle compli¬ cated by infection.1"11 Similarly, re¬ current detachments develop occa¬ sionally in buckling procedures with¬ out an encircling element to maintain the initial buckle height. Macular Hole
Macular holes account for less than 1% of all retinal detachments.114 Preop¬ erative macular cysts or holes were observed in 6% of the detachments in this series. The vast majority of pre-
macular holes do not con¬ tribute to the presence of subretinal fluid.13 In this series only a single persistent detachment could be attrib¬ uted to a macular hole. The detach¬ ment resolved with penetrating dia¬ thermic treatment of the macula.
operative
Inadequate Chorioretinal
Reaction
An insufficient chorioretinal reac¬ tion from cryopexy was implicated in only one failure following a mechani¬ cally satisfactory scierai buckling op¬ eration. Demarcation Procedures
A retinal detachment is treated most effectively by a scierai buckling procedure, resulting in closure of the retinal breaks and total reabsorption of the subretinal fluid. However, there are some cases in which this is not possible or warranted. It must be remembered that no form of retinal surgery is without some risk to the eye or the patient. Okun and Cibis14 showed that photocoagulation alone can be an acceptable alternative to scierai buckling in selected cases of limited retinal detachment. In this series of 863 primary detachments, a demarcation procedure was chosen on 79 occasions (9%). Patients treated in this manner usually had small detach¬ ments anterior to the equator, fre¬ quently had spontaneous demarcation lines, or were in poor general health or refused conventional surgery. Demar¬ cations consisted of barricading the detachment against the ora serrata with retinal cryopexy (65%), argon laser photocoagulation (24%), or xenon arc photocoagulation (11%). There were no significant differences in the success rates among the three modali¬ ties. The demarcation was regarded as successful in preventing extension of the detachment in 57 cases (72%). This is not significantly different from the 76% success rate achieved initially on all primary cases (Table 1). Twentytwo cases subsequently required addi¬ tional surgical procedures. In 17 patients the retinal detachment ex¬ tended beyond the surgical demarca¬ tion line without obvious cause. These cases were assigned arbitrarily to the category of inadequate chorioretinal reaction (Tabulation). Symptomatic small detachments should be regarded as unstable and capable of relatively rapid progression, which can exceed the rate of formation of a strong chorioretinal adhesion. New tears with small detachments outside the original area of detach¬ ment developed in two patients in this
Downloaded From: http://archopht.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/26/2015
Table 1.—Anatomie Success Rates Initial Success Rate, % 76
No. of Cases 863
Operation No.
190* 27
"Forty-eight
additional
No. of Cases
Detachment Final Success Rate, % 89
Reoperated 142
72 52 75 cases
Following Retinal
27
entered the
Significance Control
of Failures
After Retinal Detachment William F.
•
Rachal, MD, Thomas C. Burton,
One thousand eighty-eight
consecu¬
tive
operations for retinal detachment were analyzed to determine the influence of current methods of examination and treatment on failures following surgery. The majority of failures were produced by preretinal membrane formation (33%) and massive preretinal retraction (27%). Other causes of failure included undetected reti¬ nal breaks (13%), inadequate scierai buckle (10%), new retinal breaks (8%), inadequate chorioretinal reaction (7%), and iatrogenic retinal breaks (2%). Pri¬ mary operations yielded an initial cure rate of 76%. Successful reoperations raised the final cure rate to 89%. Reoper¬ ations were associated more frequently with preretinal membrane formation and had a lower final cure rate (83%). This study emphasizes the importance of recognition and management of preretinal membrane formation. Except for new reti¬ nal tears and massive preretinal retrac¬ tion, surgical failures can be avoided by improved utilization of current examina¬ tion and operative techniques.
(Arch Ophthalmol 97:480-483, 1979)
few decades ago failures after "^ retinal detachment surgery often were attributed to high myopia, apha¬ kia, extent of the detachment, and inflammation.1 Inability to localize retinal breaks was not regarded as noteworthy. Gradually there has de¬ veloped the realization that an inade¬ quate surgical technique results in the Accepted
for publication June 8, 1978. From the Department of Ophthalmology, University of Iowa Hospitals and Clinics, Iowa
City. Reprint requests to C. S. O'Brien Library, Department of Ophthalmology, University of Iowa Hospitals and Clinics, Iowa City, IA 52242 (Dr Burton).
Surgery
MD
majority of failures in retinal surgery. In 1964 Okamura et al2 emphasized that the fundamental problem in
unsuccessful retinal surgery is the inability to close all the retinal tears contributing to the detachment. In addition, there has been an increasing respect for proliferative tissue on or beneath the retina that frequently prevents closure of retinal tears. With the popularization of wider scierai buckling materials and the development of improved examina¬ tion techniques with the binocular indirect ophthalmoscope, scierai in¬ dentation, and the peripheral fundus
lens, surgical success rates improved steadily. This study
contact
have
undertaken to determine the influence of current methods of exam¬ ination and treatment on the causes of failure following retinal detachment was
surgery. SUBJECTS AND METHODS One thousand eighty-eight consecutive cases of rhegmatogenous retinal detach¬ ment operated on at the University of Iowa
Hospitals and Clinics between January April 1973 were evaluated. Not included in the study were 205 patients with retinal breaks without a surrounding 1969 and
detachment. A retinal detachment was defined as at least 1 disc diameter of fluid surrounding the margins of a retinal break. Surgical failure was defined as the continued presence of subretinal fluid that was not confined to the peripheral fundus or was not stable. Surgical techniques consisted of lamellar scierai dissection with buried solid silicone rubber implants (61%), solid silicone rubber expiants (12%), silicone sponge rubber expiants (7%), demarcation with photoco¬ agulation or cryopexy (9%), combinations of the preceding operations (10%), and
other procedures (1%). Cryopexy was used in all but 11 cases of scierai buckling oper¬ ations. Eighty-two percent of the opera¬ tions used an encircling silicone band. Data on drainage of subretinal fluid were not available. There were 863 primary cases; 190 cases required a total of 225 additional operative procedures. One hundred fifty-six patients underwent two or more scierai buckling procedures. Twelve patients had a scierai buckle followed by argon laser or xenon photocoagulation to contain residual sub¬ retinal fluid. Eighteen patients had a primary demarcation procedure by cryocoagulation or photocoagulation followed by a scierai buckling operation. Four addi¬ tional patients were treated by multiple attempts at demarcation without scierai
buckling. Photocoagulation
to increase the chorio¬ retinal reaction on the buckle during the immediate postoperative period was con¬ sidered part of the primary operation. Sixty-four cases treated in this manner were not considered as reoperations.
RESULTS
The primary cause for each of the 259 failures following the initial reti¬ nal detachment procedures is tabu¬ lated as follows. The causes of failure of subsequent operations are not included since preretinal membrane formation becomes disproportionately more common with multiple reopera¬ tions. Cause of Failure
No. of Cases
(%)
Massive preretinal retraction 69(26.6) Preretinal membrane 85(32.8) Undetected retinal tear 34(l3.1) Inadequate scierai buckle 25(9.7) New retinal tear 20(7.7) Inadequate chorioretinal reaction 18(6.9) Iatrogenic retinal tear 5(1.9) Loss of buckle height 2(0.8) Macular hole 1(0.4) Total
Downloaded From: http://archopht.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/26/2015
259(99.9)
Inadequate closure of known retinal breaks was by far the most common cause of failure of the primary proce¬ dures, accounting for 77% of all primary failures and 69% of reoperations. Massive Preretinal Retraction
Massive preretinal retraction (MPR) is regarded as synonymous for the older term "massive vitreous retraction" (MVR) and the newer term "massive periretinal prolifer¬ ation" (MPP).-'4-"' In this situation extensive proliferative tissue, proba¬ bly derived from retinal glial and pigment epithelial cells, results in a "morning glory" detachment, which is inoperable by current methods of scierai buckling, or reduces visual function to light perception.57 Massive preretinal retraction devel¬ oped in 69 of 863 primary cases (8%), which were regarded subsequently as inoperable. The risk of MPR rises to after 13.8% two operations (P
.05).
Preretinal Membrane Formation
Preretinal
membrane
formation
histologically is probably identical to the proliferative tissue that produces inoperable MPR. Clinically the mem¬ brane is expressed in less severe form and is recognized by the appearance of rolled-edge tears, star folds, equa¬ torial ridges, and fixed radial folds. In this series preretinal membrane for¬ mation was observed preoperatively in 68% of patients undergoing reoperations, which was significantly higher than the 28% rate observed in primary cases (P < .01). Surgical failure was not ascribed to preretinal membrane formation un¬ less the choice and position of implants were technically satisfactory and no other contributing factors =
could be determined. Fishmouth tears, stiff radial folds communicating with posterior fluid, and tears that reopen¬ ed on the buckle without loss of buckle height were included in this catego¬ ryUndetected Retinal Breaks
Retinal breaks that are not ob¬ served preoperatively may be grouped in three categories: (1) missed retinal breaks that were present preopera¬ tively; (2) iatrogenic breaks produced at surgery; and (3) new retinal tears
arising postoperatively. Iatrogenic breaks occur most com¬ monly during drainage of subretinal fluid, occasionally at sites of anchor¬ ing sutures for encircling elements, or
while performing lamellar dissection in thin sclera. Presentation of formed vitreous almost invariably indicates the production of a new retinal break. An exception to this may occur if a drainage site is selected beneath a large retinal break or a giant tear through which vitreous can enter the sclerotomy. Usually the surgeon is aware of such a complication and can perform immediate remedial proce¬ dures to prevent subsequent reoperations. Conversely, the retina can be perforated by a deep penetration of a needle or diathermic tip without vitreous loss. This complication, espe¬ cially if posterior to the scierai buckle, is the usual cause of reoperations from iatrogenic breaks. A retinal break was considered new if the area in which it occurred were adequately examined preoperatively, if it were completely attached in the early postoperative phase, or if the tears were large or multiple and unlikely to have been missed preoper¬
atively.
All other tears that were not observed before the first operation considered missed retinal were breaks. Failures were attributed to missed retinal breaks regardless of any other contributing factor, includ¬ ing preretinal membrane formation.
Inadequate Scierai Buckle The position or size of the scierai buckle was implicated in any case in which the entire retinal break was not isolated on the apex or plateau of the buckle and fluid around the break communicated with the residual reti¬ nal detachment. In addition, the choice of implant was implicated if it were believed to be of insufficient height or width to compensate for the amount of preretinal membrane for¬ mation known to exist preoperative¬
ly.8·9
Loss of Buckle
Height
Loss of buckle height may produce new tears or reopen previously sealed tears. A typical example of this prob¬ lem is recurrent detachment following removal of a scierai buckle compli¬ cated by infection.1"11 Similarly, re¬ current detachments develop occa¬ sionally in buckling procedures with¬ out an encircling element to maintain the initial buckle height. Macular Hole
Macular holes account for less than 1% of all retinal detachments.114 Preop¬ erative macular cysts or holes were observed in 6% of the detachments in this series. The vast majority of pre-
macular holes do not con¬ tribute to the presence of subretinal fluid.13 In this series only a single persistent detachment could be attrib¬ uted to a macular hole. The detach¬ ment resolved with penetrating dia¬ thermic treatment of the macula.
operative
Inadequate Chorioretinal
Reaction
An insufficient chorioretinal reac¬ tion from cryopexy was implicated in only one failure following a mechani¬ cally satisfactory scierai buckling op¬ eration. Demarcation Procedures
A retinal detachment is treated most effectively by a scierai buckling procedure, resulting in closure of the retinal breaks and total reabsorption of the subretinal fluid. However, there are some cases in which this is not possible or warranted. It must be remembered that no form of retinal surgery is without some risk to the eye or the patient. Okun and Cibis14 showed that photocoagulation alone can be an acceptable alternative to scierai buckling in selected cases of limited retinal detachment. In this series of 863 primary detachments, a demarcation procedure was chosen on 79 occasions (9%). Patients treated in this manner usually had small detach¬ ments anterior to the equator, fre¬ quently had spontaneous demarcation lines, or were in poor general health or refused conventional surgery. Demar¬ cations consisted of barricading the detachment against the ora serrata with retinal cryopexy (65%), argon laser photocoagulation (24%), or xenon arc photocoagulation (11%). There were no significant differences in the success rates among the three modali¬ ties. The demarcation was regarded as successful in preventing extension of the detachment in 57 cases (72%). This is not significantly different from the 76% success rate achieved initially on all primary cases (Table 1). Twentytwo cases subsequently required addi¬ tional surgical procedures. In 17 patients the retinal detachment ex¬ tended beyond the surgical demarca¬ tion line without obvious cause. These cases were assigned arbitrarily to the category of inadequate chorioretinal reaction (Tabulation). Symptomatic small detachments should be regarded as unstable and capable of relatively rapid progression, which can exceed the rate of formation of a strong chorioretinal adhesion. New tears with small detachments outside the original area of detach¬ ment developed in two patients in this
Downloaded From: http://archopht.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/26/2015
Table 1.—Anatomie Success Rates Initial Success Rate, % 76
No. of Cases 863
Operation No.
190* 27
"Forty-eight
additional
No. of Cases
Detachment Final Success Rate, % 89
Reoperated 142
72 52 75 cases
Following Retinal
27
entered the
Significance Control
