High-resolution spectral-domain OCT for PCO characterization
We used high-resolution spectral-domain OCT images to show the characteristics of different PCO types (e.g., see Figure
1). OCT facilitates the high-resolution cross-sectional imaging of the tissue [
20,
21] and was thus used for additional PCO characterization. With our OCT method, we were able to distinguish different types of PCO and to measure the IOL/PC distance in two dimensions. Until now ultrasound biomicroscopy (UBM) [
22‐
24] and Scheimpflug imaging [
25,
26] have mostly been used for the evaluation of PCO, late-postoperative CBDS and IOL positioning in the capsular bag. OCT has rarely been used to analyze a PC or PCO [
20,
27‐
29]. These few studies mostly employed OCT with a lower resolution. In this context, we present high-resolution spectral-domain OCT as a tool for future studies of changes to the PC following an Nd:YAG laser capsulotomy procedure.
It has been demonstrated previously [
30,
31] that OCT has sufficient reproducibility and can also be used to analyze anterior segment features [
20,
32‐
34], including the PC [
20,
27‐
29,
35,
36]. Moreno-Montañes et al. [
29] presented OCT with a lower resolution as a method for the objective evaluation of PCO. Using OCT, these authors estimated the PCO intensity and thickness in patients after cataract surgery. Although they showed good and acceptable interoperator repeatability for measuring posterior capsule thickness, there is still no commercially available application for PCO characterization using OCT. To obtain the IOL/PC distance map, automatic software would substantially help to reduce the amount of time needed to obtain the type of results presented in our study. Proper software would be useful for a three-dimensional analysis of different PCO types, PCO development in different IOLs, and studies of the IOL position after cataract surgery. Because the OCT has no observer bias [
29] and allows a more objective analysis than slit-lamp images, the OCT method also has the potential to become an additional tool in grading systems for PCO. However, for these purposes, automatic image processing is required to speed up the image analysis and to allow the measurement of the IOL/PC distance over the selected area (not only a single-line scan).
OCT characteristics of different PCO types
The literature refers mainly to two basic PCO types: the fibrosis and pearl types. These two basic morphological types are characterized by cell types that are responsible for their development [
37]. The results of our OCT measurements of IOL/PC distances show that the distance distribution in these two basic PCO types is practically uniform, i.e., it is practically independent of the axial position.In clinical practice, in some cases, the slit-lamp examination shows the characteristics of both basic PCO types. For this reason, we included mixed-type PCO, as a type where pearl-type and fibrosis-type PCOs are found within the same PCO. The mixed-type PCO can be easily recognized from the OCT images by the irregular IOL/PC distance (Figure
1, fourth row). In this case, the distance distribution (the third graph from the top in Figure
4) varies from the values that are typical for the pearl type (the left-hand side of the graph) to the values that are typical for the fibrosis type (the right-hand side of the graph).
During our procedures, we observed in some patients a gap between the IOL and the PC that was filled with a transparent or milky fluid. These cases correspond to late-postoperative CBDS, described by Miyake et al. [
10]. The amount of fluid was different for the various cases: in some cases, the accumulated fluid pushed the PC posteriorly and the gap was obvious; in other cases, the eyes disclosed only a small amount of fluid [
38]. Using a slit-lamp, most of these cases combined the morphological characteristics of a fibrosis-type PCO and late-postoperative CBDS. When the IOL/PC distance was large enough, a gap between the IOL and the PC was visible under the slit-lamp. Otherwise, the gap became more obvious during the procedure, when a surge of fluid flowed into the vitreous. The gap was also clearly visible on the OCT images, which were sometimes required for the analysis before the procedure to find differences with the “ordinary” fibrosis-type PCO. Our observations match cases described by Miyake et al. [
38], where late-postoperative CBDS is shown in conjunction with other types of PCO. Because we expect that the IOL/PC gap influences the total-pulse energy required for the Nd:YAG capsulotomy, we recommend using OCT imaging in clinical studies as an additional tool for PCO characterization.
Our OCT method can be used to distinguish late-postoperative CBDS from other types of PCO. While late-postoperative CBDS is a clinical diagnosis, OCT imaging could be useful for precisely documenting the presence of the condition. Our observations are in agreement with those of Tan et al. [
39], also based on analysis of case series, where OCT is recommended in the diagnosis of late-postoperative CBDS when the turbid, whitish fluid and the distance between the IOL and PC are not obvious.
To characterize the PCO types based on the IOL/PC distances, we also added a control group composed of patients without PCO after an uneventful cataract surgery. In this case, the PC is in close contact with the IOL, and the median IOL/PC distance is 11.5 μm (minimum 8.5 μm, maximum 14.8 μm) (see Table
1). This value can serve as an estimate of the thickness of the PC after the cataract surgery. Additionally, we compared the IOL/PC distances of the fibrosis-type PCO with the IOL/PC distances measured in the control group (cg and f in Figure
5, and the data listed in Table
1). The difference was statistically significant (
p = 0.02). However, in the borderline cases, the fibrosis-type PCO could not be recognized only from the OCT image, and a slit-lamp examination was still required.The most common PCO types, i.e., the fibrosis and pearl types, differ not only in the IOL/PC distance but also in how they behave during the Nd:YAG laser procedure. The difference is visible in Figure
2, which shows the OCT, slit-lamp and retroillumination images immediately after the Nd:YAG capsulotomy of the mixed-type PCO. The OCT image in Figure
2 (left) reveals the difference of the PC opening for different PCO types: on the left-hand side a pearl-type PCO is predominant, while the fibrosis type exists on the right-hand side of the PC. The pearl-type PCO mostly retracts and rolls under the IOL, while the fibrosis-type PCO, which consists of myofibroblast-like cells, appears more rigid. Thus, it does not allow the PC to retract away from the laser-induced breakdown site; instead, when treating with laser pulses, the PC acts more like a door on a hinge.
Our results reveal that the IOL/PC distances differ significantly between PCO types. This finding is in agreement with the results presented by Moreno-Montañes et al. [
29].
Influence of the PCO Type and IOL/PC distance on the total-pulse energy
The main goal of our study was to find the influence of
(i) the PCO types and
(ii) the IOL/PC distance measured by the OCT system on the total-pulse energy required to create a posterior capsulotomy of a certain area. The relationship between the PCO types and the total-pulse energy per area is presented in Figure
6, and it is statistically significant. Moreover, from the result presented in Figure
7, it can be concluded that the IOL/PC distance also significantly affects the total-pulse energy needed for the capsulotomy. Here, the estimated regression coefficient implies that for a 100-μm increase in the IOL/PC distance, we can expect a decrease of 3.6 mJ/mm
2 of the total-pulse energy per area. However, the fit of the line is not perfect; in particular, some cases with small distances require a large amount of energy. The small sample size (47 eyes of 40 patients) does not allow us to find a more complex model.
Results of our study indicate that the distance between the PC and IOL plays an important role in the usage of the total-pulse energy during Nd:YAG capsulotomy. We believe that lower total-pulse energy at larger distances is the result of the dynamics of the secondary phenomena (a cavitation bubble and a shockwave) of laser-induced breakdown near different surfaces. Previous
in-vitro studies [
40‐
42] observed the cavitation bubble and the shockwave near elastic membrane or solid surface. However, according to the results presented in this paper, it would be worth observing these phenomena also in an
in-vitro experiment where the elastic membrane will be moved some distance away from the solid surface. Thus, the presented results should help to build an
in-vitro experimental setup, which will be able to provide new insights into photodisruptive mechanisms responsible for the PC cut at different distances from the IOL. Such
in-vitro results could be clinically applicable for designing the laser system with optimal backfocus for different types of PCO.
Limitations of the study
IOL material and design play an important role in the development of PCO, as was shown in a number of previous studies [
28,
43‐
46]. Although we documented the IOL model implanted in our patients and took care that the control group had a similar distribution of IOL models to those of the other groups, the groups of patients with the same IOL model were not large enough to provide any statistically significant conclusions about how and if the material and design of the IOL affect the use of energy during Nd:YAG capsulotomy. However, we believe that the influence of IOL material and design on the total-pulse energy required for Nd:YAG capsulotomy deserves a detailed statistical analysis of larger groups of patients with the same IOL model in the future studies.
The horizontal (
x) and vertical (
y) dimensions of capsulotomies were measured using the slit-lamp light beam. We approximated the capsulotomy area with rhombus having the diagonals in
x and
y axes. However, in the further studies we suggest to calculate this area more accurately by image processing, e.g., as was performed in the study of Kanellopoulos et al. [
47].
We also could not create a non-masking design of the study. This limitation could introduce bias from the operator towards the certain type of PCO and the total energy used. To minimize operator bias and to ensure a non-biased study, we strictly defined the study protocol as described in the Methods section.
Due to the similar dependence of the total-pulse energy on (i) the PCO type and (ii) the IOL/PC distance, we cannot claim that the IOL/PC distance is the only factor affecting the total-pulse energy that is needed to perform a capsulotomy. During our procedures we also found that the repeatability of the laser focus position (with respect to the PC capsule) depends on the patient’s cooperation, the surgeon’s skill, the angle of the capsulotomy contact glass and the quality of the laser system. The influence of the aforementioned working conditions cannot be neglected. For these reasons, we believe that it is worth performing an in-vitro experiment, under more reproducible conditions, that takes into account the findings of the present study. We believe that such an in-vitro experiment will clarify the isolated influence of the IOL/PC distance on the total-pulse energy and may reveal the main mechanisms that are responsible for the PC opening.