Download The Long-term Effects of Laser Photocoagulation Treatment in

The Long-term Effects of Laser
Photocoagulation Treatment in Patients
with Diabetic Retinopathy
The Early Treatment Diabetic Retinopathy Follow-up
Study
Emily Y. Chew, MD,1 Frederick L. Ferris III, MD,1 Karl G. Csaky, MD, PhD,2 Robert P. Murphy, MD,1
Elvira Agrón, MSc,1 Darby J. S. Thompson, MSc,3 George F. Reed, PhD,1 Andrew P. Schachat, MD4
Objectives: To evaluate the long-term natural history and effects of laser photocoagulation treatment in
patients with diabetic retinopathy.
Design: Follow-up study of the 214 surviving patients enrolled originally at the Johns Hopkins Clinical Center
for the Early Treatment Diabetic Retinopathy Study (ETDRS), which was a clinical trial designed to evaluate the
role of laser photocoagulation and aspirin treatment in patients with diabetic retinopathy.
Methods: Early Treatment Diabetic Retinopathy Study patients enrolled in the Johns Hopkins Clinical Center
had complete eye examinations, including best-corrected visual acuity measurements, fundus photographs, and
medical questionnaires throughout the 7-year study. They had the same examinations at the final long-term
follow-up visit at the National Eye Institute, National Institutes of Health, 13 to 19.5 years after the initial laser
photocoagulation (median, 16.7 years).
Main Outcome Measures: The major outcomes were mortality and the rates of moderate and severe vision
loss. The secondary outcomes were progression of diabetic retinopathy and need for other eye surgery.
Results: Of the 214 patients who were alive at the end of the original ETDRS in 1989, 130 (61%) were
deceased at the time of the re-examination. Of the 84 who were alive, 71 (85%) were examined at their long-term
follow-up visit at the National Institutes of Health. At the long-term follow-up examination, 42% had visual acuity
of 20/20 or better, and 84% had visual acuity of 20/40 or better in the better eye. Compared with baseline, 20%
of patients had moderate vision loss (loss of 3 lines or more vision) in the better eye at follow-up. Only one patient
had visual acuity of 20/200 bilaterally. He had visual acuity loss secondary to age-related macular degeneration.
No patient had severe vision loss (worse than 5/200). All the initially untreated eyes of patients who had severe
nonproliferative diabetic retinopathy or worse by the time of the ETDRS closeout visit of the original study
received scatter photocoagulation treatment. Focal photocoagulation was performed in 43% bilaterally and 22%
unilaterally. Cataract surgery was performed in 31% of the patients, vitrectomy in 17%, and glaucoma surgery
in one patient.
Conclusions: As previously reported, the mortality rate of patients with diabetic retinopathy is much higher
than that of the general population. For those who survived, aggressive follow-up, with treatment when indicated,
seems to be associated with maintenance of good long-term visual acuity for most patients. The need for laser
scatter photocoagulation with long-term follow-up seems to be high. Ophthalmology 2003;110:1683–1689 ©
2003 by the American Academy of Ophthalmology.
Diabetic retinopathy is a leading cause of vision impairment
in the adult population in the United States.1 The treatment
Originally received: August 21, 2002.
Accepted: February 19, 2003.
Manuscript no. 220573.
1
National Eye Institute/National Institutes of Health, Division of Epidemiology and Clinical Research, Bethesda, Maryland.
2
National Eye Institute/National Institutes of Health, Laboratory of Immunology, Bethesda, Maryland.
3
EMMES Corporation, Rockville, Maryland.
© 2003 by the American Academy of Ophthalmology
Published by Elsevier Inc.
strategies for diabetic retinopathy developed over the last
several decades are, in part, based on National Institutes of
Health–supported clinical trials evaluating diabetes control,
4
Wilmer Eye Institute, Johns Hopkins Medical Institutions, Baltimore,
Maryland.
Reprint requests to Emily Y. Chew, MD, Division of Biometry and
Epidemiology, National Eye Institute/National Institutes of Health, Building 31, Room 6A52, 31 Center Drive, MSC-2510, Bethesda, MD 208922510.
ISSN 0161-6420/03/$–see front matter
doi:10.1016/S0161-6420(03)00579-7
1683
Ophthalmology Volume 110, Number 9, September 2003
laser photocoagulation, and vitrectomy.1–5 The efficacy of
laser photocoagulation for proliferative diabetic retinopathy
was demonstrated in the Diabetic Retinopathy Study
(DRS).3 Treatment with photocoagulation, either laser or
xenon arc, reduced the rate of severe vision loss (visual
acuity of worse than 5/200 for at least 4 months’ duration)
by 50%. This treatment became the standard of care for
patients with proliferative diabetic retinopathy.
In the subsequent clinical trial of diabetic retinopathy,
the Early Treatment Diabetic Retinopathy Study (ETDRS),
3711 patients were randomly assigned to early laser photocoagulation and aspirin treatment and followed for 5 to 9
years. Patients had retinopathy that ranged from mild nonproliferative retinopathy to early proliferative retinopathy.
The data from the ETDRS suggested that scatter laser
photocoagulation should be considered for all eyes with
severe nonproliferative diabetic retinopathy or worse and
not for those with mild to moderate nonproliferative retinopathy.5 For patients with type 2 diabetes, early treatment
at the severe nonproliferative stage rather than waiting for
the onset of high-risk proliferative retinopathy is considered, because the rate of severe loss was reduced by more
than 50% in such eyes treated with early laser photocoagulation compared with eyes assigned to deferral of laser
photocoagulation.6 This beneficial effect of early treatment
at the severe nonproliferative stage was not seen in patients
with type 1 diabetes. Many factors need to be considered in
deciding when to initiate photocoagulation for these patients, but it is apparent from both the DRS and ETDRS that
virtually all patients with high-risk proliferative retinopathy
should have scatter photocoagulation without delay.
Focal laser photocoagulation treatment for diabetic macular edema was also evaluated in the ETDRS. The treatment
of eyes with clinically significant macular edema (defined as
retinal edema that affects or threatens the center of the
macula) with focal laser photocoagulation reduced the risk
of moderate vision loss (loss of 3 or more lines on a
logarithm of the minimum angle of resolution visual acuity
chart) by 50%.4 Focal treatment with laser photocoagulation
has become the standard treatment for diabetic macular
edema.
The results of the Diabetic Retinopathy Vitrectomy
Study, a randomized clinical trial of vitrectomy for eyes
with dense vitreous hemorrhage and for eyes with very
severe proliferative diabetic retinopathy, showed a greater
chance of recovering good vision with vitrectomy.7 With
the improvement in vitrectomy techniques and instrumentation, the visual results have markedly improved since
these early trials. Vitrectomy has become an important part
of the treatment strategy for patients with diabetic retinopathy that is not resolved by photocoagulation.
A review of the effects of using all these therapeutic
strategies has demonstrated that their timely implementation
can reduce the risk of severe vision loss by as much as
90%.8 Only 4% of eyes and 1% of patients with proliferative diabetic retinopathy in the ETDRS had severe vision
loss at 5 years. This is remarkably lower than the 50%
5-year blindness rate seen in patients before the availability
of photocoagulation and also estimated from untreated eyes
of the patients enrolled in the DRS and other early studies of
1684
diabetic retinopathy.3,9 –11 Although these treatment strategies have become the standard therapy, their long-term
effects have not been assessed beyond 10 years.
Patients enrolled in the original ETDRS at the Johns
Hopkins Clinical Center were re-examined to determine the
long-term rates of severe and moderate vision loss and
further ocular complications requiring further surgery. This
information on the long-term effects of treatment is important in providing care for all patients affected with diabetes,
particularly with the increasing incidence of type 2 diabetes
in the United States. The improvement in medical care of
patients with diabetes has resulted in increased survival.
This growing population of patients with diabetes has increased the public health importance of the treatment of
diabetes and its complications.
Materials and Methods
The ETDRS was designed to assess photocoagulation and aspirin
treatment for patients with mild nonproliferative to early proliferative diabetic retinopathy.12 Study patients were randomly assigned to 650 mg of aspirin or placebo daily. One eye of each
patient was randomly assigned to early laser scatter and/or focal
photocoagulation, whereas the fellow eye was given scatter laser
photocoagulation only when high-risk proliferative diabetic retinopathy developed. At baseline, these patients were aged 18 to 69
years with a favorable prognosis for 5-year survival.
All patients who were enrolled and followed in the original
ETDRS at the Johns Hopkins Medical Institution and who were
living at the closeout visit of August 1989 were identified (n ⫽
214). Those still living were invited for an eye examination at the
National Eye Institute, National Institutes of Health. The institutional review boards for research on human subjects at both the
John Hopkins University and the National Eye Institute/National
Institutes of Health approved the follow-up study. A total of 3711
patients had been enrolled in 22 ETDRS clinical centers and
followed from 5 to 9 years. This follow-up study was conducted
from 1997 to 2000 on the patients enrolled at the Johns Hopkins
Clinical Center, providing 13 to 19.5 years of follow-up (median,
16.7 years) after the study randomization visit and the initiation of
laser photocoagulation.
The surviving patients were all invited for an examination at
the clinical center at the National Eye Institute/National Institutes
of Health. After obtaining signed informed consents, historical
information was obtained. Patients received a comprehensive eye
examination, which included best-corrected visual acuity, slitlamp biomicroscopy, and dilated ophthalmoscopy. Lens opacities
were assessed clinically by slit-lamp biomicroscopy, using the
Age-Related Eye Disease Study classification system.13 Stereoscopic fundus photography of seven fields and fluorescein angiography were performed. The Fundus Photograph Reading Center at
the University of Wisconsin, in Madison, centrally graded and
compared the fundus photographs and fluorescein angiograms
obtained at this follow-up visit with those obtained at baseline and
during the course of the study.
Six patients were willing to be examined but were unable to
travel to the National Eye Institute for examination. These patients
were examined by collaborating ophthalmologists, most of whom
were coinvestigators in the original ETDRS study. They examined
these patients in their institutions and private offices using the
standardized protocol for examination and fundus photography
and completing the data collection forms used at the National Eye
Institute.
Chew et al 䡠 Effects of Laser Photocoagulation in Diabetic Retinopathy
Table 1. Baseline Characteristics
Johns
Hopkins
N (%)
Total
Age
⬍20
20–29
30–39
40–49
50–59
60–69
Gender
Male
Female
Race
White
Black
Hispanic
Pacific Islander
American Indian
Diabetes type
Type 1
Mixed
Type 2
Duration of diabetes
⬍10
10–19
ⱖ20
Insulin use
No
Occasionally
Daily
Smoking
Never
Stopped ⱖ2 yrs
Stopped ⬍2 yrs
Currently smokes
Systolic blood pressure (mmHg)
ⱕ120
121–149
150⫹
Diastolic blood pressure (mmHg)
ⱕ76
77–89
90⫹
Antihypertensive medications
Yes
No
History of coronary artery disease
Yes
No
Suspect
History of myocardial infarction
Yes
No
Suspect
History of congestive heart failure
Yes
No
Suspect
History of stroke
Yes
No
Suspect
214 (100)
Table 1. (continued)
All Other
Early
Treatment
of Diabetic
Retinopathy
Study
Clinics
P
N (%)
Value
3497 (100)
3 (1.4)
49 (1.4)
38 (17.7) 536 (15.3)
26 (12.1) 536 (15.3)
35 (16.3) 571 (16.3)
64 (29.9) 1005 (28.7)
48 (22.4) 800 (22.8)
0.817
120 (56.0) 1976 (56.5)
94 (43.9) 1521 (43.4)
0.902
174 (81.3) 2660 (76.0)
40
463 (13.2)
0
337 (9.6)
0
24 (6)
0
13 (3)
⬍0.0001
55 (25.7) 1075 (30.7)
81 (37.8) 1348 (38.5)
78 (36.4) 1074 (30.7)
0.146
33 (15.4) 577 (16.4)
138 (64.4) 1981 (56.6)
43 (20.0) 939 (26.8)
0.055
47 (21.9) 546 (15.6)
1 (4)
4 (1)
166 (77.5) 2947 (84.2)
0.013
104 (48.5) 1536 (43.9)
35 (16.3) 834 (23.8)
17 (7.9)
152 (4.3)
58 (27.1) 974 (27.8)
0.009
54 (25.2) 887 (25.4)
104 (48.6) 1497 (42.8)
56 (26.2) 1112 (31.8)
0.165
46 (21.5) 876 (26.1)
90 (42.1) 1637 (46.8)
78 (36.5) 983 (28.1)
0.032
41 (19.1) 618 (17.6)
173 (80.8) 2878 (82.3)
0.582
18 (8.4)
173 (4.9)
192 (89.7) 3232 (92.4)
4 (1.8)
91 (2.6)
0.071
16 (7.4)
146 (4.1)
195 (91.1) 3305 (94.5)
3 (1.4)
45 (1.2)
0.071
11 (5.1)
63 (1.8)
200 (93.4) 3405 (97.3)
3 (1.4)
28 (8)
0.004
4 (1.8)
42 (1.2)
208 (97.1) 3473 (98.3)
2 (9)
17 (4)
0.294
All Other
Early
Treatment
of Diabetic
Retinopathy
Johns
Study
Hopkins
Clinics
P
N (%)
N (%)
Value
History of transient ischemic attack
Yes
No
Suspect
Proteinuria
Negative, trace
1⫹, 2⫹, 3⫹, 4⫹
Cholesterol
⬍200
200–240
⬎240
Mean
Median
Range
Serum creatinine (mg/dl)
⬍1.0
1.0–1.3
⬎1.3
Mean
Median
Range
Hemoglobin A1c (%)
⬍10
10⫹
Mean
Median
Range
Ocular characteristics
Total
Visual acuity at baseline
20/20 or better
⬍20/20 to ⱖ20/40
⬍20/40 to ⱖ20/200
⬍20/200
Severity of retinopathy at baseline
Mild to moderate NPDR
Severe NPDR
PDR
Severity of retinopathy at last visit
Mild to moderate NPDR
Severe NPDR
PDR
Macular edema at baseline
Yes
No
Intraocular pressure at baseline
(mmHg)
⬍14
14–20
⬎20
Photocoagulation given during the
original study
Scatter
Yes
No
Focal
Yes
No
2 (9)
11 (3)
209 (97.6) 3446 (98.5)
3 (1.4)
39 (1.1)
0.210
157 (73.3) 2498 (71.5)
57 (26.6) 992 (28.4)
0.573
47 (25.9) 860 (34.0)
54 (29.8) 783 (30.9)
80 (44.1) 885 (35.0)
239⫾4
228⫾1
235
220
138–426
106–852
0.025
0.014
80 (38.0) 1236 (37.2)
109 (51.9) 1724 (52.0)
21 (10.0) 354 (10.6)
1.05⫾0.02 1.07⫾0.01
1.00
1.00
0.5–2.1
0.1–5.6
0.942
102 (56.3) 1449 (58.1)
79 (43.6) 1043 (41.8)
9.85⫾0.15 9.67⫾0.04
9.73
9.62
3.30
3.03
0.637
0.232
0.290
214 (100) 3497 (100)
122 (57.0) 1936 (55.3)
73 (34.1) 1238 (35.4)
19 (8.8) 311 (8.8)
0 (0)
11 (3)
0.834
83 (38.7) 1432 (40.9)
92 (42.9) 1438 (41.1)
39 (18.2) 627 (17.9)
0.814
80 (38.4) 1218 (36.3)
27 (12.9) 475 (14.1)
101 (48.5) 1660 (49.5)
0.788
144 (67.2) 2388 (68.2)
70 (32.7) 1109 (31.7)
0.761
35 (16.4) 695 (19.9) ⬍0.0001
149 (69.6) 2588 (74.2)
30 (14.0) 207 (5.9)
101 (47.1) 1366 (39.0)
113 (52.8) 2131 (60.9)
0.018
110 (51.4) 1518 (43.4)
104 (48.5) 1979 (56.5)
0.022
NPDR ⫽ nonproliferative diabetic retinopathy; PDR ⫽ proliferative diabetic retinopathy.
1685
Ophthalmology Volume 110, Number 9, September 2003
Table 2. Factors Associated with Increased Mortality
Variable
Hazard
Ratio
95%
Hazard
Confidence
Ratio
Limits
P
Value
Age (yrs)
Hemoglobin A1c (%)
Proteinuria (mg/dl)
1.067
1.153
2.268
1.050
1.044
1.515
1.085
1.274
3.394
⬍0.0001
0.0049
⬍0.0001
The names and demographic data of all patients who did not
respond to the invitation for a follow-up examination were submitted to the National Death Index. Matches from the National
Death Index provided confirmation of the deceased patients. The
causes of death were obtained from the death certificates.
Statistical Methods
Comparisons of categorical baseline characteristics between the
Johns Hopkins Clinic patients and the ETDRS patients enrolled in
the remaining clinical centers were conducted by chi-square test
analysis or by permutation testing of the chi-square test statistic if
the cell sizes were too small to warrant application of the large
sample method. Continuous baseline variables were compared by
the t test. Cox proportional hazards regression was applied to
assess and adjust for the effects of the following risk factors on
mortality in the follow-up group: glycosylated hemoglobin
(HBA1c), age, duration of diabetes, proteinuria, blood pressure,
visual acuity score, and diabetic retinopathy severity. Because
these are secondary analyses, the P value considered to be statistically significant is P⬍0.01.
of the ETDRS population, one sees that the patients were similar
in age and gender. There were more blacks but fewer Hispanic
patients in the Johns Hopkins clinic population than in the general
ETDRS population. Patients enrolled in the general ETDRS population were more likely to be cigarette smokers than in the Johns
Hopkins clinic population. There were more patients in the Johns
Hopkins clinic that had history of congestive heart disease compared with the ETDRS population. Other medical and baseline
characteristics were similar in both the Johns Hopkins clinic population and the remaining ETDRS population. The levels of severity of retinopathy at baseline were also similar in the two
populations, whereas patients from the Johns Hopkins Clinical
Center had somewhat higher intraocular pressures at baseline.
Visual Acuity
At the ETDRS baseline examination, the visual acuities of the
better eye of the 71 patients who participated in this long-term
follow-up study were 20/20 or better in 57%, 20/40 or better in
91%, and less than 20/40 to 20/100 in 9%. At the ETDRS Follow-up Study examination, best-corrected visual acuities in the
patients’ better eye were as follows: 20/20 or better in 42%, 20/40
or better in 84%, between 20/40 and 20/100 in 15%, and 20/200 in
1 patient (1%). The decreased visual acuity in this patient was
secondary to geographic atrophy associated with age-related macular degeneration. There were no cases of severe vision loss (visual
acuity of worse than 5/200). The rate of moderate vision loss (loss
of 15 letters or more at follow-up compared with baseline visual
acuity) was 20%.
Ocular Surgeries
Results
Mortality
Of the 214 patients who were alive at the end of the original
ETDRS study in 1989, 130 (61%) were deceased by the time of the
follow-up examination. The causes of death were determined by
assessment of the death certificates obtained from the National
Death Index report. These consist of cardiovascular disease in
48%, nonspecific diabetic complications in 28%, renal disease in
4%, neurologic causes in 7%, malignancies in 4%, and other
causes in 5%. The adjusted analyses (Cox proportional hazards
model) of baseline risk factors associated with death for this clinic
population included increasing age, proteinuria, and elevated
HBA1c (Table 1). In univariate analyses, with or without age
adjustment, both poor visual acuity and more severe retinopathy
were significantly associated with increased mortality. However,
in the final Cox model, these two factors were no longer statistically significant, but the sample size of this study is fairly small.
Of the 84 who were still living, 1 patient was hospitalized for
Alzheimer’s disease, 1 was confined to a nursing home, 1 had
incomplete data from the patient’s ophthalmologist, and 10 could
not be contacted. Excluding these 13 patients, 71 (85%) were
examined. The duration of follow-up from the beginning of the
enrollment in the original ETDRS until the follow-examination
was 13 to 15 years for 16 patients (23%), 16 to 17 years for 25
patients (35%), and more than 17 years for 30 patients (42%), with
median follow-up of 16.7 years. Table 2 shows a comparison of
the baseline characteristics of the patients enrolled in the Johns
Hopkins Clinic with the study population enrolled in the remaining
ETDRS clinics nationwide.
Comparing the patients enrolled at Johns Hopkins with the rest
1686
Photocoagulation and Vitrectomy. As stipulated by the study
protocol, all patients enrolled in the ETDRS received laser photocoagulation in one eye immediately after being randomly assigned
to their groups. At the closeout ETDRS visit of the 71 patients in
1989, 47% and 40% of the eyes assigned to deferral of laser
photocoagulation had received focal and scatter photocoagulation,
respectively. By the time of the long-term follow-up examination,
all of the initially untreated eyes of patients who had severe
nonproliferative diabetic retinopathy or worse had eventually required and received scatter photocoagulation treatment. Ten of the
14 patients who had mild to moderate nonproliferative retinopathy
at the closeout visit of the original study had not progressed
beyond moderate nonproliferative retinopathy at the long-term
follow-up visit and had not received scatter photocoagulation. The
other 4 patients in this group had received scatter photocoagulation
at some time between the end of the ETDRS and the time of the
long-term follow-up visit. Other interventions noted at the time of
the long-term follow-up visit were as follows: focal photocoagulation had been performed in 43% of patients bilaterally and 22%
unilaterally, and vitrectomies had been performed in 12 (17%)
patients and glaucoma surgery in 1 patient.
Lens Opacities
Of the patients examined at the follow-up study, cataract surgery
had occurred in 22 (31%) patients; 12 were bilateral and 10 were
unilateral. Of the phakic patients, 25% had posterior subcapsular
cataract greater than 1 mm in diameter, 35% had cortical opacities
occupying at least 12% of the lens area, and 30% had nuclear lens
opacities greater than Age-Related Eye Disease Study standard
photograph number 4.13 Seventy-four percent had one or more of
these lenticular opacities.
Chew et al 䡠 Effects of Laser Photocoagulation in Diabetic Retinopathy
Figure 1. The 5-year rates of severe vision loss of Early Treatment of
Diabetic Retinopathy Study patients with proliferative diabetic retinopathy were 1% in patients and 4% in eyes. This is remarkably lower than the
50% 5-year rate seen in the untreated eyes of patients enrolled in the
Diabetic Retinopathy Study.
Figure 2. The mortality of the Early Treatment of Diabetic Retinopathy
Study patients is compared with that of a general population, adjusted for
age.
Discussion
Visual Acuity Results
More than half of the patients originally enrolled in the
ETDRS clinic at Johns Hopkins were deceased at the time
of the long-term follow-up examination, and those patients
with the worst visual acuity at the ETDRS closeout visit
were the least likely to survive. Most of the patients who
survived maintained fairly good visual acuity at the time of
the long-term follow-up examination; 42% of patients in
this study retained visual acuity of 20/20 or better, and 84%
retained visual acuity of 20/40 or better in at least one eye.
These data suggest that those patients who receive photocoagulation and who survive long-term are more likely to
have maintained good visual acuity during the early phases
of treatment. They are also likely to maintain fairly good
visual acuity with long-term follow-up. Although they continue to have diabetes, for most, the retinopathy seems to
eventually become quiescent. This is due, at least in part, to
the beneficial effects of laser photocoagulation and vitrectomy seen at 5 years.8 The short-term evaluation of these
therapeutic strategies has demonstrated highly beneficial
effects in reducing the risk of severe vision loss by more
than 90% in 5 years. Only 4% of eyes and 1% of patients
with proliferative diabetic retinopathy in the ETDRS had
severe vision loss at 5 years. This is remarkably lower than
the greater than 50% 5-year blindness rate seen in patients
before the availability of photocoagulation and also estimated from untreated eyes of the patients enrolled in the
DRS (Fig 1).3
The 10-year visual acuity results of the 51 patients originally enrolled in DRS at the Bascom Palmer Eye Institute
were reported in 1991. Of these patients, 24 had visual
acuity of 20/40 or better (47%), and 7 were worse than
20/200 (14%).14 The comparison of the visual acuity results
in this ETDRS Follow-up Study with these visual acuities
and the acuities of the untreated eyes before the availability
of photocoagulation reflect an improvement in the treatment
of the diabetes, as well as the development and refinement
of techniques such as laser photocoagulation and vitrectomy. It is also important to note that most of the ETDRS
patients required further scatter laser photocoagulation during follow-up after the closeout visit, underscoring the need
for vigilant follow-up during the lifetime of patients with
diabetic retinopathy.
Mortality Results
Patients with the worst visual acuity at the end of the
original ETDRS had particularly high mortality rates. Although the visual acuity results reported here might reflect
the beneficial effect of laser and other medical treatments,
there is also a selection bias such that those patients with
better visual acuity were more likely to survive and to be
available for the long-term follow-up examination. At the
time of the ETDRS closeout visit in 1989, 8.4% of the
Hopkins patients were legally blind, and 4.7% had severe
visual loss in at least one eye. No patient with legal blindness at closeout survived to be examined in the follow-up
study. More than 75% of patients who survived during the
follow-up for the examination at the National Eye Institute
had closeout visual acuity of 20/20.
The high rate of mortality of the ETDRS Follow-up
Study is reflected in the comparison with that of a general
population, adjusted for age (Fig 2). The deaths occurred
over the course of the follow-up, as shown in the survival
curve (Fig 3). In the evaluation of the mortality among
adults with or without diabetes in the National Health and
Nutrition Examination Survey, the 22-year mortality for
those with diabetes was 33.7% for persons aged 25 to 54
years of age at baseline and 82.9% for those aged 55 to 75
years compared with 10.2% and 64.4% for those people
without diabetes, respectively.15 The rates of mortality were
markedly higher for persons with diabetes compared with
the rates of those without diabetes. The risk of death was
doubled in the National Health and Nutrition Examination
Survey for those patients who had definite proteinuria at
1687
Ophthalmology Volume 110, Number 9, September 2003
Figure 3. The deaths that occurred over the course of the study follow-up
are shown in this survival curve. The death rates are adjusted for baseline
visual acuity.
baseline compared with those without proteinuria at baseline. The risk factors associated with mortality in the ETDRS follow-up population included baseline characteristics
of age, proteinuria, and elevated HBA1c. The hazard ratio
for an increase of age from 45 to 75 years was 7.0. The
hazard ratio for an increase of HBA1c from 7% to 12% was
2.0 and from 9% to 12% was 1.5.
The possible association of baseline risk factors of visual
acuity and severity of retinopathy with mortality was investigated using the entire ETDRS data set for the duration of
the original study (data not shown). Both decreased baseline
visual acuity and increased severity of the baseline level of
retinopathy were found to be associated with increased
mortality in the univariate analyses. However, when these
factors were entered into a multivariate analysis, they were
no longer statistically significant. Other factors such as
proteinuria, presence of cardiovascular disease at baseline,
and elevated serum cholesterol levels were found to be
associated with increased mortality in the multivariate analyses. In these analyses, it is important to note that the ocular
risk factors might not be considered traditional risk factors
of mortality but might be so-called risk markers, concomitant diabetic complications that might be markers of systemic disease. For example, decreased baseline visual acuity
and increased severity of retinopathy might be indicative of
generalized vascular disease that might lead to increased
mortality. They might explain in part the relatively good
visual acuity in the survivors examined in the follow-up
study.
In the analyses of a population-based study, the Wisconsin Epidemiologic Study of Diabetic Retinopathy showed
that at 6 years of follow-up, the ocular risk factors associated with increasing mortality were poorer visual acuity and
more severe retinopathy in both types 1 and 2 diabetes.16
Again, it is possible that poorer visual acuity and more
severe retinopathy at baseline might be markers for poor
systemic vascular status of the patient. Other ocular risk
factors associated with decreased survival in the Wisconsin
Epidemiologic Study of Diabetic Retinopathy included
1688
glaucoma in the younger population and cataract in the
older population.
A limitation to the current ETDRS Follow-up Study is
the relatively small number of patients who were examined.
However, the comparison of the baseline characteristics of
the Johns Hopkins clinic with the entire ETDRS population
showed that this particular clinic is representative of the
entire cohort. A larger study might improve the precision of
the results, but given the dramatic findings, it seems reasonable to assume that the direction and the size of the
estimates would not change in a clinically important way. In
addition, these findings seem to be in concert with the
clinical impression of the retinal specialists regarding the
prognosis of quiescent diabetic retinopathy.
In summary, the long-term visual acuity results, after
treatment strategies that include laser photocoagulation and
vitrectomy in those survivors in the ETDRS Follow-up
Study, are remarkable, with 82% of patients retaining driving vision in at least one eye. These results emphasize the
need for vigilant follow-up throughout the lifetime of patients with diabetic retinopathy, because further scatter
treatment was required in most patients followed. Although
laser photocoagulation is a highly effective treatment for
diabetic retinopathy, diabetic retinopathy remains a major
cause of visual impairment in adults in the United States.
Further improvement of screening methods and treatment
delivery for patients with diabetes is needed. Both patients
and physicians need to be aware of the beneficial effects of
tight glycemic, serum lipid, and blood pressure control on
diabetic retinopathy, as well as the long-term benefits of
photocoagulation and vitrectomy that are currently available
for those patients who have clinically important retinopathy.
References
1. Klein R, Klein BEK, Moss SE. Visual impairment in diabetes.
Ophthalmology 1984;91:1–9.
2. The Diabetes Control and Complications Trial Research
Group. The effect of intensive treatment of diabetes on the
development and progression of long-term complications in
insulin-dependent diabetes mellitus. N Engl J Med 1993;329:
977– 86.
3. The Diabetic Retinopathy Study Research Group. Photocoagulation treatment of proliferative diabetic retinopathy. Clinical
application of Diabetic Retinopathy Study (DRS) findings,
DRS report number 8. Ophthalmology 1981;88:583– 600.
4. Early Treatment Diabetic Retinopathy Study Research Group.
Photocoagulation for diabetic macular edema: Early Treatment Diabetic Retinopathy Study report number 1. Arch Ophthalmol 1985;103:1796 – 806.
5. Early Treatment Diabetic Retinopathy Study Research Group.
Early photocoagulation for diabetic retinopathy: ETDRS report number 9. Ophthalmology 1991;98(5 suppl):767– 85.
6. Ferris F III. Early photocoagulation in patients with either type
I or type II diabetes. Trans Am Ophthalmol Soc 1996;94:505–
37.
7. The Diabetic Retinopathy Vitrectomy Study Research Group.
Early vitrectomy for severe vitreous hemorrhage in diabetic
retinopathy: two-year results of a randomized trial. Diabetic
Retinopathy Vitrectomy Study Report 2. Arch Ophthalmol
1985;103:1644 –52.
Chew et al 䡠 Effects of Laser Photocoagulation in Diabetic Retinopathy
8. Ferris FL III. How effective are treatments for diabetic retinopathy? JAMA 1993;269:1290 –1.
9. Beetham WP. Visual prognosis of proliferating diabetic retinopathy. Br J Ophthalmol 1963;47:611–9.
10. Caird FI, Burditt AF, Draper GJ. Diabetic retinopathy: a
further study of prognosis for vision. Diabetes 1968;17:121–3.
11. Deckert T, Simonsen SE, Poulsen JE. Prognosis of proliferative
retinopathy in juvenile diabetes. Diabetes 1967;10:728 –33.
12. Early Treatment Diabetic Retinopathy Study Research Group.
Early Treatment Diabetic Retinopathy Study design and baseline patient characteristics. ETDRS report number 7. Ophthalmology 1991;98(5 suppl):741–56.
13. The Age-Related Eye Disease Study Research Group. The
Age-Related Eye Disease Study (AREDS) system for classifying cataracts from photographs: AREDS Report No. 4. Am
J Ophthalmol 2001;131:167–75.
14. Blakenship GW. Fifteen-year argon laser and xenon photocoagulation results of Bascom Palmer Eye Institute’s patients
participating in the Diabetic Retinopathy Study. Ophthalmology 1991;98:125– 8.
15. Gu K, Cowie CC, Harris MI. Mortality in adults with or
without diabetes in a national cohort of the U.S. population,
1971–1993. Diabetes Care 1998;21:1138 – 45.
16. Klein R, Moss SE, Klein BE, DeMets DL. Relation of ocular
and systemic factors to survival in diabetes. Arch Intern Med
1989;149:266 –72.
1689