Académique Documents
Professionnel Documents
Culture Documents
PURPOSE: To determine whether intraocular glistenings have an impact on light scatter and visual
function.
SETTING: John A. Moran Eye Center Laboratories, University of Utah, Salt Lake City, Utah, USA.
DESIGN: Retrospective cohort study.
METHODS: Pseudophakic patients with visual acuity no worse than 0.02 logMAR and no ocular pa-
thology were enrolled. All had received a single-piece high-refractive-index acrylic intraocular lens
(IOL). All IOLs were photographed, and glistenings were analyzed for size and density. Outcome
measures included logMAR corrected distance visual acuity (CDVA), mesopic 10% contrast
logMAR CDVA with and without glare, and straylight determination with a straylight meter
(C Quant log).
RESULTS: All 79 patients had glistenings within 2 diameter groups: 6 to 25 mm and over 25 mm.
Linear regression for the nonstratified group was significant for IOL glistening size versus contrast
visual acuity with glare. Linear regression for the 6 to 25 mm group was significant for a measure of
severity index (%area) versus the straylight meter measurements, %area/size versus straylight me-
ter measurements, IOL age versus CDVA, IOL age versus contrast visual acuity, and IOL age versus
contrast visual acuity with glare. Linear regression for the over 25 mm group was significant for IOL
age versus glistening size and %area/size versus contrast visual acuity, and density versus CDVA
and contrast visual acuity with glare.
CONCLUSIONS: Glistening %area, at a key size, correlated with random light scatter. The age of the
IOL likely affects glistening size and visual parameters.
Financial Disclosure: No author has a financial or proprietary interest in any material or method
mentioned.
J Cataract Refract Surg 2015; 41:1190–1198 Q 2015 ASCRS and ESCRS
Glistenings are fluid-filled microvacuoles within an The relationship between glistening formation and
intraocular lens (IOL). They form when several, if time since surgery has also been a subject of consider-
not all, types of hydrophobic IOLs are exposed to an able debate. Multiple studies report progression of
aqueous environment.1–3 The effect of IOL glistenings glistening formation with time,6,7 while others report
on visual function has been an area of considerable a lack of progression 1 year after implantation.8,9 By
controversy. DeHoog and Doraiswamy4 recently evaluating our cohort of patients whose IOLs varied
calculated the effect of glistening size and density on in age between 1 year and 4 years, we would have
light scatter measured by the modulation transfer the opportunity to further clarify this controversy as
function and concluded that for the same percentage well.
of the IOL blocked by glistenings, a smaller size would The impact of glistenings on visual function has been
have a greater effect on image degradation. In our pre- documented previously, with correlations observed
vious study of pseudophakic dysphotopsia,5 we pho- between glistenings and high-frequency contrast
tographed all IOLs and realized that this study could acuity loss,7,10,11 visual acuity,8,9 and significant visual
provide clinical support or refutation of DeHoog and impairment to the point of IOL explantation.1,12 In
Doraiswamy’s4 theoretical findings. contrast, other reports show no significant impact of
(rather than 50 years, which was the lower age limit of the density (%area/size) that was similar to the analysis
cohort in the Kinard et al. study5). described by DeHoog and Doraiswamy.4 These measure-
All eligible patients who consented to participate had an ments were compared versus CDVA, contrast visual acuity
ophthalmic examination that included slitlamp bio- with and without glare, and straylight meter log to look at
microscopy and imaging, with and without retroillumina- straylight scatter and visual quality. Linear correlation coef-
tion, to confirm the presence of glistenings. Further testing ficients were used to examine relationships between IOL age
included photopic logMAR CDVA after manifest refraction, and %area, glistening size, and %area/size as well as rela-
10% contrast logMAR acuity (Regan charts, Good-Lite Co.) tionships between visual parameters including CDVA,
with mesopic lighting (13 ft/4 m ETDRS charts with contrast visual acuity with and without glare, and the log
photopic lighting at 85 candelas [cd]/m2 and mesopic light- of straylight. Statistical analysis was completed using Stat-
ing at 3 cd/m2) with and without glare. Glare of car lights at plus:mac software (version 2009, Analystsoft, Inc.). A
night was simulated with 2 halogen beams, 1 on each side of P value less than 0.05 was considered statistically significant.
the low-contrast acuity chart. Straylight scattering was
measured using the C-Quant straylight meter (Oculus
Optikger€ate GmbH). RESULTS
Glistening quantification and analysis methods were The records of 2953 patients were reviewed, and 79 (9
similar to those reported by Waite et al.7 Multiple boxes more patients than the cohort in the Kinard et al.
ranging from 0.5 to 1.0 mm2 in area within the visual axis
from the clearest single image of each IOL were analyzed, study) met all criteria and enrolled. Table 1 shows
and measurements were averaged. Glistenings were evalu- the characteristics for the nonstratified data. The
ated using the public domain Java image-processing pro- SN60WF IOL was implanted in 36 eyes (45.5%), the
gram ImageJ software.17 SN60AT in 36 eyes (45.5%), the SN60T5 in 4 eyes
Intraocular lenses were placed in 2 relatively equal groups (5.2%), the SN60T4 in 1 eye (1.3%), the SN60T3 in 1
based on average glistening size (diameter) of 6 to 25 mm or
greater than 25 mm. For each size grouping, a linear correla- eye (1.3%), and the SN6AD3 in 1 eye (1.3%).
tion coefficient was computed with a measure of severity in- Table 2 shows the characteristics of the 2 glistening
dex (%area) described by Waite et al.,7 size, and a measure of diameter groups (6 to 25 mm and O25 mm). The
diameter ranged from 6.0 to 36.0 mm (2.8E-5 to 1.0E-3 (R Z 0.37, P Z .014) (Figure 3, bottom). Although not
mm2 in area), and the count/box area ranged from 2 statistically significant, there was also a relationship
to 336 glistenings/mm2. with %area versus CDVA, contrast visual acuity, and
Linear regressions for the entire group (nonstrati- contrast visual acuity with glare (CDVA: R Z 0.16,
fied) were significant for positive correlations between P Z .346; contrast visual acuity: R Z 0.29, P Z .09;
IOL age and glistening size (R Z 0.26, P Z .019) contrast visual acuity with glare: R Z 0.21, P Z
(Figure 1, left) as well as between IOL age and contrast 0.226) (Figure 4).
visual acuity with glare (R Z 0.28, P Z .014) (Figure 1, Linear regression for the over 25 mm diameter group
right). They approached significance for correlation be- was significant for a positive correlation between
tween IOL age and CDVA (R Z 0.22, P Z .050) IOL age and glistening size (R Z 0.45, P Z .006)
(Figure 1, bottom). (Figure 5, top), and there was a negative correlation
Linear regressions for the 6 to 25 mm diameter group between %area/size and contrast visual acuity
were significant for positive correlations between (R Z 0.34, P Z .048) (Figure 5, bottom). Negative corre-
%area and straylight meter measurements (R Z 0.32, lations were observed between glistening density
P Z .033) (Figure 2, left), and %area/size and stray- (counts per box) and CDVA (R Z 0.39, P Z .02) and
light meter measurements (R Z 0.30, P Z .047) contrast visual acuity with glare (R Z 0.34, P Z .04),
(Figure 2, right). A negative correlation for size and and approached significance with contrast visual
straylight meter measurements also approached sig- acuity (R Z 0.33, P Z .05) (Figure 6).
nificance (R Z 0.29, P Z .052) (Figure 2, bottom). Signif-
icant positive correlations were also found for IOL age
versus CDVA (R Z 0.033, P Z .031) (Figure 3, left), IOL DISCUSSION
age versus contrast visual acuity (R Z 0.33, P Z .031) Glistenings of variable size and density were found in
(Figure 3, right), and contrast visual acuity with glare 100% of the IOLs of patients enrolled in this study. Our
findings are similar to previously reported diameters created more scatter even at an equal percentage
(5.2 to 30.7 mm) and count/box area measurements of obscured area.
(0.8 to 223 glistenings/mm2) in a study that used Previously reported measurements of straylight
similar quantification methods.7 Although the clinical scatter using the C-Quant correlated with glare.18–20
impact of glistenings remains to be determined defin- It is not surprising that glistenings would affect stray-
itively, the literature consistently has suggested that light and, consistent with the DeHoog and Doraisw-
visual function can be affected by glistening forma- amy report,4 that the impact is greatest in a range of
tion.1,4,7–12 smaller glistening sizes. These findings are particu-
The results of the DeHoog and Doraiswamy theo- larly interesting because our straylight meter log read-
retical study4 suggest that at the same %area, a ings were clearly important in a specific small
smaller size would create more light scatter and vi- glistening group, even though the actual straylight
sual quality degradation. Our cohort, although meter values were low (little scatter) compared with
small, naturally split at above and below 25 mm. the results in other reports.21,22
Figure 7 shows slitlamp photographs (original In our study, we did not measure glistenings less
magnification 25) of IOL glistenings with average than 6 mm in diameter, so we might not be able to
glistening sizes in the 6 to 25 mm group and in the detect smaller sizes with our photographs. The De-
over 25 mm group. Furthermore, our evaluation Hoog and Doraiswamy study4 showed that 2 mm glis-
confirmed DeHoog and Doraiswamy’s findings in tenings had the most impact on light scatter;
2 ways: (1) Light scatter as shown in the C-Quant however, 2 mm glistenings might not be detectable
straylight meter was only significantly correlated even with the slitlamp, and it seems unlikely that
with the less than 25 mm size. (2) In the less than we could detect glistenings of this size. Thus,
25 mm group there was a significant linear correla- although our study cannot prove or disprove effects
tion between %area/size and the straylight meter of smaller glistenings on light scatter or visual func-
measurements, showing that smaller glistenings tion, it is possible that glistenings not visible at the
slitlamp could have clinical significance. Further Using the correlation coefficient for this relationship
studies that are focused on detecting even the small- in our cohort, a post hoc power calculation indicated
est of glistenings with the slitlamp are needed to that enrollment of approximately 200 patients would
address this issue. have been necessary for an association to reach signif-
In our cohort, the average glistening size increased icance. Because increasing size decreases scatter, this
with IOL age. These findings are in line with previous might be an advantage to increasing IOL age;
reports of glistenings that worsened over time.6,7 however, a concurrent increase in the %area would
Although the %area did not increase with time, this likely increase scatter. The overall clinical impact of
could be a result of our sample size (N Z 79), which increasing glistening size and the %area would likely
might have been inadequate to detect a difference. depend on which of these parameters changed the
Figure 5. Correlations in the more than 25 mm group. Left: The IOL age correlation with size. Right: The %area/size correlation with contrast
visual acuity (IOL Z intraocular lens; VA Z visual acuity).
most over time. The relative impact of changes in size In future studies of intraocular glistenings, the addition
compared with changes in the %area might be better of surface light scattering as an endpoint, as well as spe-
addressed with a larger study with sufficient power cial attention to and documentation of smaller vacuole
to examine these relationships in greater depth. glistenings (!25 mm), will help to further elucidate the
The age of the IOL was also correlated with contrast role of glistening size and surface light scattering on
visual acuity with glare and approached significance IOL opacification and visual functioning. In addition,
with CDVA. These findings were also seen in the less the impact of increasing IOL age on visual function
than 25 mm group, along with correlations between might also include external factors such as age-related
IOL age and contrast visual acuity without glare. How- decline and although we have attempted to control for
ever, correlations were not seen between IOL age and vi- these factors with thorough examinations and tight
sual parameters in the more than 25 mm group, again screening to exclude any other type of pathology, it is
suggesting that IOLs with smaller glistenings had a possible such factors contributed to these findings.
greater impact on visual function than IOLs with larger As was the case with the Kinard et al. study,5 this
glistenings. In general, the %area was also greater in the cohort was super normal, with low levels of light scat-
6 to 25 mm cohort. Although these IOL age correlations ter and excellent visual quality metrics in all enrolled
might simply be the result of the glistenings alone, it is patients. At some level of light scatter, there is bound
probable that surface light scattering, which has also to be a correlation with clinical glare. Interestingly, in
been shown to increase with IOL age for the IOL studied, the 6 to 25 mm group, the %area versus CDVA, contrast
contributes to decreased visual function as well, espe- visual acuity, and contrast visual acuity with glare all
cially with glare. Although we did not include IOL opa- had similarly suggestive but nonsignificant correla-
cification resulting from surface light scatter as an tions. The similarity between all 3 of these relation-
outcome measure, previous studies1,15,23,24 suggest that ships would be unusual if there were no correlation
surface irregularity progresses over time for the IOL at all. A power calculation using these correlation coef-
studied; this might have been a factor in our findings. ficients shows that with a sample size between 85 and
194, there would be sufficient power for these associa- of our positive findings is supportive of our suggested
tions to reach significance; however, only 44 such pa- conclusions. The contrary results further point out the
tients were included in our study sample. We critical need for a well-powered, more definitive study.
observe that the issue of underpowered studies has In conclusion, we suggest that our data provide clin-
been a factor in all clinical glistening studies to date, ical support for DeHoog and Doraiswamy’s simulation
including this study. The impact of IOL glistenings findings4 that at smaller glistening sizes, a similar
merits a large and adequately powered study, both %area obscured by glistenings is associated with
to definitively answer questions and to assess changes increased light scatter. We also provide additional ev-
over time as millions of patients have and will idence that glistening size increases with IOL age and
continue to have this phenomenon. The present pilot that visual parameters also decrease with increasing
study represents the largest population studied to IOL age; possibly, this is related both to glistenings
date without confounding visual problems that might and to IOL surface changes. Larger clinical studies
affect visual quality. We hope that our study will lead are needed to further examine the effects of glistening
to a study with a larger sample and will point the way size, the %area, and the %area/size on visual function
to delineation of glistening correlates of visual quality. parameters. We also understand that the manufacturer
In the more than 25 mm group, we found 4 correlations of the IOL studied has implemented recent changes in
that run counter to our other findings. The %area/size the manufacturing process in an attempt to minimize
when compared with contrast acuity would suggest the formation of glistenings. Clinical validation and
that visual quality was better with a smaller size (all still testing of the formation of IOL glistenings over time us-
O25 mm however). Furthermore, density measure ing the different IOL manufacturing models for com-
(number of glistenings per mm2 irrespective of size) parisons of visual functioning parameters similar to
correlated negatively (the greater the density, the better those reported in our study is needed to examine the
the visual quality) with CDVA and contrast acuity with true clinical impact of such changes.
and without glare. That all were in the more than 25 mm
group (mean %area of glistenings was 40% less than in
the 6 to 25 mm group) and that the correlation was
WHAT WAS KNOWN
only for density (density did not show a significant cor-
relation in the 6 to 25 mm group) suggest these are false- Glistenings in hydrophobic acrylic IOLs have been docu-
positive findings. Nonetheless, these correlations exist mented. Conflicting reports have evaluated the clinical
and further support the case for a significantly powered significance of glistenings and the specific impact on vi-
definitive study that focuses on the specific findings that sual acuity, glare, and light scatter.
our pilot study suggests.
Weaknesses of this study are the relatively small WHAT THIS PAPER ADDS
number of patients, the lack of severe manifestations
of the glistening phenomenon, and that any patient The %area and %area/size of glistenings correlated with
with less than 20/20 1 CDVA or who had ever com- light scatter at a key size of 6 to 25 mm. These findings
plained about their visual quality were also excluded. suggest that at a specific size, glistenings have an effect
Because of the difficulty with recruitment of patients ful- on light scatter.
filling all inclusion criteria, a relatively small number of The IOL age also had negative correlations with CDVA,
patients were enrolled in each group (44 patients in the 6 contrast visual acuity, and contrast visual acuity with
to 25 mm group; 35 patients in the O25 mm group). As glare, highlighting the impact of IOL age on visual function
the number of comparisons increases, so does the risk in patients with 6 to 25 mm diameter glistenings.
for false-positive correlations. However, the consistency