Secondary intraocular lens implantation: a large retrospective analysis

Ideally, after uneventful cataract surgery, a posterior chamber IOL (PC-IOL) is implanted in the capsular bag. However, this is not always possible, as capsular bag-associated complications may already exist preoperatively (loose zonula, IOL luxation) or occur intraoperatively (anterior or posterior capsular tear). In these cases, either no IOL will be implanted (aphakia) or the IOL has to be fixated in other positions such as anterior chamber (AC), iris, sulcus, or the sclera.

In cases of secondary aphakia or IOL-related complications, a secondary intraocular lens implantation is the preferable surgical procedure. IOL luxation, incorrect IOL power, IOL opacification, uveitis-glaucoma-hyphema (UGH) syndrome, patient dissatisfaction, or secondary aphakia indicate the major reasons for such surgery. Secondary IOL implantations have increased over recent years, and this surgical procedure is now considered common [1, 2].

This retrospective longitudinal case series was conducted in order to identify preoperative ocular risk factors and indications for secondary IOL surgery in a tertiary vitreoretinal referral center and compare the postoperative complications, refractive and visual outcome of such surgery. A secondary analysis examined the influence of preoperative ocular risk factors, biometry, surgeon, IOL fixation, and postoperative complications on the refractive and visual outcome. All secondary intraocular lenses were placed in one of the following positions: anterior chamber (AC, angle supported), retropupillary iris-claw fixation (IC-IOL), sulcus without optic capture, capsular bag, or trans-scleral fixation.

Five-hundred seventy-five eyes underwent a secondary IOL implantation over the examined period of time (2009–2014). However, only a total of 182 eyes (98 right, 84 left) of 174 patients (100 male, 74 female) with a mean age of 63 years old (range 15–92 years old) were evaluated and analyzed in this study, due to lack of follow-up of other patients. Eight patients underwent surgery on both eyes.

The most often reported preoperative risk factor was previous complete pars plana vitrectomy (78 eyes, 43%). Eight of these eyes had, in addition, previous retinal surgery. Overall, previous retinal surgery was noted as the second most frequent preoperative ocular risk factor (62 eyes, 34%), followed by trauma (30 eyes, 17%), PXF (22 eyes, 12%), and uveitis (16 eyes, 9%). Myopia was observed in more than 35% of the cases. Moderate myopia, with an axial length (AL) between 28 and 30 mm, was observed in 10 cases, and high myopia, with an AL > 30 mm, in 7 cases.

Figure 1 demonstrates the frequency of each indication for secondary IOL implantation. Leading cause for surgery was IOL dislocation. Specifically, an in-the-bag IOL dislocation was observed in 112 of 134 eyes (84%) and an out-of-the bag IOL dislocation in 22 eyes (16%). Moreover, 47 of 134 (35%) eyes had an IOL dislocation beneath the optical zone, resulting to a high ammetropia of the eye preoperatively.

The mean follow-up time was 17 ± 13.6 months (range 3–60 months). The mean time period between the primary surgery and the secondary IOL implantation was 8.4 ± 6.5 years (range 0–32 years).

Four experienced surgeons performed all the surgical procedures. Surgeon 1 performed 39%, surgeon 2 23%, surgeon 3 20%, and surgeon 4 18% of the surgeries. A scleral incision of different lengths was always performed at the superior part of the eye (position between 70°-120°). During the observation time (2009-2013), the most frequent final position of the implanted IOL was the anterior chamber (angle supported AC-IOL, 92 eyes, 50% of all cases). A retropupillary fixation of an iris-claw IOL (IC-IOL, Verisyse®) was performed in 22% of the cases (40 eyes). A sulcus fixation was available in 15% of the cases (28 eyes), and an in-the-bag fixation was performed in 6% of the cases (10 eyes). A scleral fixation was performed in 7% of the cases (12 eyes). For the IOL power calculation, the biometric formula SRK-T was used in 67% of the cases, followed by the Haigis in 17% and the Hoffer Q in 7%. No biometric formula was used in 9% of the cases, due to re-fixation of the primary implanted IOL.

While epidemiologic data referring to the incidence of IOL dislocation shows an incidence of 0.1–3% of implanted IOL after 5 years, most experts agree that the general incidence of secondary IOL implantation due to different reasons is increasing [2, 4].

IOL dislocation, aphakia (from various causes), IOL opacifications, patient dissatisfaction after multifocal intraocular lens implantation, or UGH syndrome constitute the major reasons for surgery [5]. This study was conducted in order to examine the incidence of preoperative ocular risk factors and indications for surgery in a large European vitreoretinal referral center, and to compare the complications rate and the refractive and visual outcome.

Different parameters were examined for their incidence as preoperative risk factors. While PXF is known to be a main risk factor for zonular instability and IOL dislocation with a presence of over 30% [5‐7], recent data reveals that PXF is frequently clinically underdiagnosed in cases of late in-the-bag IOL dislocation [8, 9]. Although in our study IOL dislocation was the major reason for surgery, PXF was present only in 12% of all cases, in 21 eyes with an IOL dislocation (16%) and specifically in 18 of 112 eyes with an in-the-bag IOL dislocation (16%). Regarding the low incidence of clinically detected PXF in our cases of dislocated IOL as compared to other studies, we cannot rule out that the incidence of PXF may have been underreported.

Further factors like trauma, previous retinal surgery, vitrectomy, uveitis, myopia, and aging are also reported in the literature [2, 10‐13] as risk factors. Interestingly, in this study, previous complete pars plana vitrectomy was the most frequent preoperative ocular risk factor for in-the-bag IOL dislocation compared to other studies [10, 14]. With increasing rates of vitreoretinal surgery, this might also reflect a factor in increasing incidence of IOL dislocation overall. On the other hand, this study was conducted in a referral center for vitreoretinal surgery, so these data might be influenced by the specific patient population.

The population’s average age in this cohort was almost a decade younger than is reported in most studies [9, 15]. This might reflect a specific patient population in this series. On the other hand, the incidence of previous trauma or uveitis was higher than in other published data [2, 5, 12, 16]. In our series, about 25% had either a previous trauma or a history of uveitis, representing a relative high incidence of these risk factors in comparison to other studies [5, 16]. The high incidence of previous vitrectomy, previous retinal surgery, previous trauma, uveitis, and the younger age reflect the complexity of the cases of this cohort, and the need for sufficient refractive results. Intraocular lens dislocation has been noted as the major reason for secondary IOL implantations [16]. In this study, the IOL dislocation was also the leading cause for surgery, with the majority of the eyes having an in-the-bag IOL dislocation. Further reasons were secondary aphakia or IOL opacifications. Eleven aphakic eyes had a previous trauma, whereas the remaining 25 eyes (69%) were aphakic due to other causes. Although the spectrum for surgery indication is quite wide, IOL dislocation and aphakia after complicated cataract surgery or other complicated intraocular surgery still remain the major reasons for secondary IOL implantation in most other recent studies as well [5, 16]. The mean time of 8.7 years between secondary IOL implantation and primary cataract surgery found in our study is in accordance with published data [5, 17].

The postoperative objective refractive spherical equivalent (ORSE) showed no statistically significant difference between the groups with regard to the fixation position of the secondary IOL (p = 0.106, Kruskal-Wallis). In contrast, refractive prediction error as well as absolute error differ statistically significantly between the different IOL positions. Postoperative ORSE and AE of this study are comparable to the published data [18, 19]. In our study, the retropupillary IC-IOLs (Verisyse®) showed the lowest AE, followed by the AC-IOLs, sulcus IOLs, and PC-IOLs, although the difference was not statistically significant. Scleral fixated IOLs were excluded from the statistical analysis due to re-fixation of the same IOL in 6 of the 12 eyes and small number of the remaining cases, but they showed comparable results to IC-IOLs, as already published in other studies [20]. In contrast to Brunin et al., sulcus fixated IOLs showed a relatively high PE and AE [16]. The fact that PE and AE show a higher deviation might be due to the fact that on an empirical basis, IOL Power was weakened upon implantation in sulcus. Thus, the intended refraction as predicted in biometry would not correspond to the targeted refraction as there is an empiric corrective factor. This assumption is supported by the fact that the use of biometric formulas did not show different results for the AE in these cases. Previous retinal surgery or trauma was noted as a risk factor in 6 of the 10 eyes with a capsular bag secondary IOL implantation, whereas a vitrectomy was performed in combination with the secondary in-the-bag IOL implantation in 8 of these 10 cases. We think that this context as well as the complexity of the cases might explain the relative high AE that was observed. Additionally, a relative high AE may have occurred due to the complexity of the preoperative optical biometry, especially in cases of IOL dislocation, where the capsular-bag/lens complex may interfere with the optical axis and lead to a further error in calculating the optimum refractive power of the IOL to be implanted.

Although preoperative CDVA in our study was worse than in other published literature, an improvement of CDVA was observed in all groups, except of sclera fixated IOLs [16]. Interestingly, this group showed in our study the best pre- and postoperative CDVA and no change after surgery, but the small number of patients of this group and the variable results do not allow any further conclusions. Considering the complexity of the cases of this study and the existing preoperative visual impairment, mainly due to previous vitrectomy or retinal surgery, only 50.9% of the eyes had a preoperative CDVA better than 0.5 LogMAR. However, 79.4% achieved a postoperative CDVA better than 0.5 LogMAR, whereas 60% of the eyes showed an improvement of CDVA.

The frequency of the postoperative AC hemorrhage in our study was not higher than other published data and was most frequently observed after iris fixation, followed by scleral fixation. AC hemorrhage did not affect the final CDVA and postoperative AE negatively. As already reported, a small self-limiting AC hemorrhage is not likely to worsen the postoperative visual acuity [2, 21].

Additionally, in this study, a low rate of postoperative hypotony was observed, compared to Rey et al. and Dajee et al. [19, 21]. A reason for that could be a smaller incision procedure or the low incidence of uveitis cases. Despite the complexity of the cases, cystoid macular edema, elevated intraocular pressure, and IOL dislocation incidence were comparable to the published literature [5, 16, 22, 23]. However, IOL dislocation occurred only in two IC-IOLs and in one AC-IOL, indicating a relative low incidence. In contrast to that, two of the 12 scleral fixated IOLs showed a dislocation and one a decentration, reflecting a rather higher complication rate. However, all these eyes underwent simultaneously complete pars plana vitrectomy, which may have affected their postoperative stability.

Predictability of the magnitude of postoperative refraction, which is expressed by AE, showed in our study correlation with indication for surgery, fixation type, and postoperative AC hemorrhage. Moreover, this study reports a lower AE in cases with dislocated IOLs compared to aphakia or IOL opacification. The highest refractive deviations were documented for the correction of secondary aphakia. In 8 of these cases, an IOL was implanted in-the-bag, whereas in 8 other cases in sulcus. However, in these groups, a large proportion of eyes have had previous trauma or vitrectomy for various reasons. Additionally, the empiric correction of the sulcus position was not reflected in the AE.

Regarding the defined as long-term complications, IOL decentration and dislocation were the most frequently observed. Interestingly, these occurred more often in eyes with secondary IOL implantation in the sulcus or in-the-bag and are not commonly observed in AC-IOLs or IC-IOLs or sclera fixated IOLs in our study. Thus, it seems that in our study with a large portion of eyes having suffered trauma, secondary IOL as AC-IOL or IC-IOL or scleral fixation is not only safer in the long term but also shows a better predictable refractive outcome with the limitations mentioned above.

Our study has several shortcomings. Firstly, it is a retrospective study with all its limitations. With regard to this retrospective design, we believe that our results rather underreport the results as patients with unsatisfying results may tend to ask for revisions. We assume that the big number of patients that preferred to undergo further follow up examinations at their ophthalmologist rather than at the clinic were satisfied with their refractive outcome. However, the lack of follow up limits the population of this cohort and most probably leads to reporting rather the results of the most complex cases, which intend to visit big referral centers for their follow up. Secondly, several surgeons using different techniques and skills were involved, which does not correspond to a standardized procedure. Additionally, this study is not limited to one condition, but also includes complicated cases such as, for example, traumatic cases.

We cannot comment on the outcome of a specific technique from our data as the surgical procedure was chosen according to the surgeons’ experience. In our setting, scleral fixation was chosen if other options were not feasible. In general, these were complex or traumatic cases. Thus, our results concerning the scleral fixation technique rather underreport the results of this technique [16]. Additionally, this technique has been modified recently and the current technique only partly corresponds to the evaluated time period [24].

Our study analyzes a large number of cases with the need for secondary IOL implantation. While surgical standards in these rather complex cases are difficult to investigate, we believe that all of these cases represent indications for secondary IOL implantation and thus are worth reporting. While one may argue in individual cases which is the preferable technique, all of these techniques have their particular indication. This claim is underlined by our results showing no significant differences in most of these techniques with regard to visual acuity, PE, and AE.

As a conclusion, we may state that our study shows that in our setting, vitrectomy seems to be a major preoperative ocular risk factor in all eyes that underwent secondary IOL implantation surgery. Additionally, all applied techniques show satisfying refractive and visual results albeit not being comparable to primary IOL implantation in the capsular bag. Retropupillary IC-IOLs showed the best refractive results, though not statistically significantly different to the other fixation positions. However, we claim that IC-IOLs seem to be the most satisfying option of secondary IOL implantation especially in complicated cases, with history of previous vitrectomy and should be preferred. Further studies analyzing techniques with recent biometric formulas and techniques in a standardized setting are warranted.