Repeat selective laser trabeculoplasty can be effective in eyes with initial modest response.

PURPOSE: To evaluate the effectiveness of repeat selective laser trabeculoplasty (SLT) in eyes exhibiting only a modest response upon initial treatment.
MATERIALS AND METHODS: Retrospective chart review was conducted of 51 eyes that received initial 360 degree SLT (SLT1) and subsequent SLT (SLT2) from 2003-2011 at a large academic ophthalmology practice. Successful response (S) was a post-treatment 12 month mean IOP reduction ≥ 20% from baseline, while modest response (M) was <20% reduction over the same time. Chi-squared and log rank analyses were used to determine if success after SLT2 depended on having successful (S1) or modest (M1) response after SLT1.
RESULTS: IOP was significantly reduced from baseline in both SLT1 and SLT2. The proportion of eyes with S2 was not significantly different between those with initial M1 or S1 (36.67% vs. 52.38%, respectively; P = 0.26). Log rank analysis revealed no differences between M1 and S1 in determining SLT2 success (P = 0.41). This outcome was similar when the analyses were performed for the right and left eye independently.
CONCLUSION: The proportion of eyes that successfully responded to repeat SLT did not differ based upon whether the response to initial SLT was successful or modest. This raises the possibility that repeat SLT should not be excluded as an option for those eyes that have only a modest initial response.

INTRODUCTION

Selective laser trabeculoplasty (SLT) is a popular initial and adjunctive treatment for open angle glaucoma. The treatment uses a 532-nm frequency-doubled q-switched neodymium: Ttrium-aluminum-garnet laser to selectively target pigmented trabecular meshwork cells.1 Trials comparing SLT to argon laser trabeculoplasty (ALT) have shown that both laser platforms produce comparable intraocular pressure (IOP) reductions.2345 Although multiple theories exist, the exact mechanism by which laser trabeculoplasty (ALT or SLT) lowers IOP remains unknown.67 Despite uncertainty regarding mechanism of action, SLT offers some potential benefits over ALT, including decreased damage to trabecular meshwork.89 These differences between the platforms may render SLT more safely repeatable owing to lesser collateral tissue damage.8

Repeat SLT has been shown to be effective after initially successful SLT (≥20% IOP reduction from baseline).10 Less is known about the effectiveness of repeat SLT in eyes with sub-optimal response to first SLT. To evaluate the possibility that repeat SLT may have an additional effect regardless of initial response, we evaluated the efficacy of repeat SLT in eyes that exhibited a successful response (≥20% IOP reduction) compared to a modest response (<20% IOP reduction) to initial SLT.

MATERIALS AND METHODS

A retrospective review of the electronic records at a large multi-physician academic ophthalmology practice was conducted after institutional review board approval. Patients with open angle glaucoma who underwent two treatments of SLT for IOP reduction between 2003 and 2011 were eligible for the study. The decision to perform SLT was based on the clinician's assessment of the need to lower IOP in each particular glaucoma patient. There were no defined IOP thresholds or initial response thresholds for repeat treatment. The decision to repeat SLT was made by individual clinicians when further IOP reduction was needed.

Definitions of baseline IOP, post-treatment 12-month mean IOP, non-responders, successful response, and modest response were defined as follows. For SLT1, baseline IOP was defined as the mean pressure from the previous two visits prior to SLT. The timing for each patient to undergo SLT2 was at the discretion of the treating ophthalmologist, which precluded the calculation of the mean pressure from two prior visits. Therefore, baseline IOP for SLT2 was defined as the last measured IOP prior to treatment. Post-treatment 12-month mean IOP was calculated as the mean of all recorded visits over 12 months. Successful response (S) was defined as a post-treatment mean IOP reduction over 12 months ≥20% from baseline, while modest response (M) was defined as a 12-month mean IOP reduction <20% but >0% from baseline.

Inclusion and exclusion criteria were defined as follows: For inclusion only those treated with 360 degree SLT at each of the two treatment sessions were included. All included subjects had angle grades of 3 or 4 and were given non-contiguous spot treatment. Patients with a history of SLT other than to 360 degree of angle, ALT, or incisional glaucoma surgery were excluded. Non-responders, defined as those whose post-treatment 12-month mean IOP remained the same or increased compared to baseline after SLT1, were excluded.

Patients who had successful (S1) and modest (M1) response upon 1st SLT (SLT1) treatment were further analyzed to determine 2nd SLT (SLT2) responses (S2 or M2). Demographics, laser treatment parameters, and number of glaucoma medications at SLT1 and SLT2 were collected. Goldmann IOP measurements were collected at baseline, 1, 4, 6, and 12 months after each SLT treatment. The windows for data collection at 1, 4, and 6 months were ±1 month and 12 months were ±2 months. The data were analyzed for means and standard deviation. Student's t-test was used to compare the mean IOP reduction, number of glaucoma medications, number of treatment spots, and energy per spot between SLT1 and SLT2. Analysis of variance (ANOVA) was used to compare the mean IOP at each follow-up time to baseline. The Pearson Chi-squared test was used to compare the proportion of patients with S2 following M1 or S1, as well as comparing those with M2 after M1 and S1. Multivariate logistic regression including the variables age, sex, race, baseline IOP, baseline number of glaucoma medications, and total energy (number of treatment spots x energy per spot in mJ) were used to explore for predictors of success of IOP reduction following first and second SLT treatments.

A Kaplan-Meier analysis with a log rank test for comparison was performed to detect differences in IOP reduction ≥20% upon SLT2, with the curves of eyes with M1 and S1 being compared. As there were 17 patients who had both eyes included, we first analyzed the right and left eyes separately to avoid any confounding of variables. After finding that the results of the separate eyes were not significantly different, we pooled the data together. A post-hoc power analysis was conducted based on the sample size of the group being analyzed. All P values are two sided and statistical significance is defined as P ≤ 0.05. All analyses were performed with Stata®11.0 SE (Stata Corp LP, College Station, TX, USA).

RESULTS

A total of 51 eyes from 34 patients (26 right eyes and 25 left eyes) met the inclusion and exclusion criteria of this study. The median time between SLT1 and SLT2 was 23 months (Mean: 26 months; Standard Deviation: 13.06 months; Range: 7-72 months). The IOP reduction for SLT1 and SLT2 was statistically significant from baseline at all time points, with no differences in mean IOP between the two treatments [Table 1]. We estimated the power of this study to detect a clinically relevant difference of 20% reduction of mean IOP at 12 months from baseline was >90% (α=0.05, two tailed).

Table 1

IOP reduction from baseline at all time points

When comparing characteristics between SLT1 and SLT2 in Table 2, we found that the laser parameters (number of shots and energy per shot) were significantly different between SLT1 and SLT2, with an overall higher number of shots upon SLT1 (P = 0.002) and higher energy per shot upon SLT2 (P < 0.001). As we wanted to make sure that the reduction in IOP was not contributed by an increase of glaucoma medications, we found that the mean number of glaucoma medications was not different between SLT1 and SLT2 [Table 2]. The mean number of glaucoma medications at baseline prior to SLT1 and SLT2, 1.47 ± 0.85, was not statistically different compared to the number at 12 months post SLT, 1.42 ± 0.89 (P = 0.32). The post treatment 12-month mean IOP reductions for SLT1 and SLT2 were similar (3.80 and 3.61 mmHg, respectively, P = 0.68). Mean baseline IOP was slightly higher in S1 (21 ± 3) and S2 (20 ± 5) than in M1 (19 ± 3) and M2 (18 ± 2), respectively, and this was statistically significant for SLT1 (P = 0.008) but not SLT2 (P = 0.10). When comparing 12-month mean IOP reduction from baseline, S1 and S2 had significantly higher percentage declines than M1 and M2, respectively (S1 and S2: 28% and 27%, respectively; M1 and M2: 12% and 11%, respectively; P < 0.001).

Table 2

Comparison of SLT1 and SLT2 groups

Multivariate logistic regression analysis of SLT1 and SLT2 is shown in Table 3. Higher baseline IOP was the only significant predictor for SLT1 success (OR = 3.72; 95% CI 1.33-10.40; P = 0.01). Using the same multivariate analysis, no covariates significantly predicted success upon SLT2.

Table 3

Multivariate logistic regression of characteristics for ≥20% IOP reduction

Table 4 shows the proportional outcome of results in both SLT treatments. Eyes that met success criteria after SLT1 and SLT2 were: S1 = 21/51 (41%), and S2 = 22/51 (43%). In the 22 eyes that experienced S2, the same number of eyes were M1 (11 eyes) compared to S1 (11 eyes). The proportion of eyes with successful or modest response did not differ significantly regardless of outcome on first SLT treatment (P = 0.26).

Table 4

Chi-squared analysis of M2 and S2 after M1 and S1 (P=0.64)

This Chi-squared analysis was also performed for right eyes and left eyes independently. In the analysis of right eyes (N = 26), the proportion of all eyes who were M1 or S1 and then went on to become S2 after SLT2 (M1 = 7/18, S1 = 6/8) was not statistically different (P = 0.09). Likewise, in the analysis of left eyes (N = 25), the Chi-squared analysis did not yield any statistical difference in the proportion of those who became S2 (M1 = 4/12, S1 = 5/13; P = 0.79).

Another confirmatory analysis using Kaplan-Meier survival curves demonstrated no significant differences between M1 and S1 in determining the success of SLT2 (log rank P = 0.41) [Figure 1]. When stratified by each eye, the log rank analysis for the right eyes was not significantly different (P = 0.23), nor was the analysis for the left eyes (P = 0.79).

Figure 1

Kaplan-Meier survival curves comparing those with S1 and M1 at SLT2 Abbreviations: SLT2, 2nd SLT treatment; S1, Successful response (≥20% IOP reduction) upon 1st SLT; M1, Modest response (<20% IOP reduction) upon 1st SLT

DISCUSSION

This retrospective study sought to explore whether initial response to SLT is predictive of response to repeat SLT. In our study, mean IOP was lower at each follow-up time compared to baseline in both SLT1 and SLT2, indicating its efficacy.111213 Similar to other studies, our findings in SLT1 confirm that higher baseline IOP is a significant predictor of IOP reduction ≥20% from baseline.141516 A higher baseline IOP however did not predict IOP reduction upon SLT2. This unexpected finding deserves to be further examined in future studies.

Because ultrastructural studies have observed SLT causing minimal damage to the trabecular meshwork compared to ALT, it has been suggested that SLT can be successfully repeated.17 While treatment with ALT has an IOP lowering effect upon initial treatment, repeat ALT is less effective.18 The exact mechanisms of action of SLT still remain unclear. There is evidence that inflammation, a result of SLT therapy, modulates the permeability of the endothelial cells of Schlemm's canal.8917]

Our observations raise the possibility that repeat SLT can be effective in IOP reduction even if the initial treatment exhibited only a modest response. We believe this topic has not yet been explored, as common teaching in clinical practice has been to repeat SLT only in eyes showing a good initial response. By reviewing records from patients exhibiting a range of responses to initial SLT, we can evaluate whether this tradition is supported. Using the strict criterion of ≥20% IOP reduction as a successful response, initial SLT was successful for only 41% of eyes. However, for this cohort, the success of SLT2 did not depend on that of SLT1, and the proportion of eyes that had a successful response following SLT2 (S2) was not significantly different based on the initial response. Further, the success rate of SLT2 was not significantly different from that of SLT1 (43 vs. 41%, respectively, P = 0.26).

The success of repeat SLT in lowering IOP has been previously reported in a retrospective case series by Hong et al.10 In keeping with customary practice, they only included subjects who had a successful initial SLT, defined as ≥20% IOP reduction for 6 months or more. The success rate following initial SLT in their population was 50% (n = 22) at 5-8 months. The success rate of repeat SLT in eyes initially successful was 43% (n = 19) at 5-8 months. They observed no statistical differences between IOP-lowering results of SLT1 and SLT2 after 5-8 months follow up (P = 0.16). However, their study did not address efficacy of repeat SLT in those who did not meet their success criterion following initial SLT.

While the repeatability of SLT in eyes initially responsive has been an accepted practice, the possibility that repeat SLT can be effective in eyes with only a modest initial response could be explained by the hypothesis that a graded response occurs in different eyes. In other words, in eyes where the response threshold is higher, a modest initial response is followed by a more robust drop in IOP upon retreatment (a “threshold” response). Similarly, a “ceiling” response to treatment could explain why some eyes respond well initially and then demonstrate little or no additional effect with repeat treatment.

This study is retrospective, with the possibility that covariables may have biased SLT outcome. Including only those that responded who received repeat SLT, either modest or successful, may also be a source of bias, but repeat SLT is not currently indicated for non-responders upon initial SLT. Both left and right eyes were included to increase the number of eyes in the study. This may have overestimated the effect of SLT on responsive patients. However, separate right and left eye analysis for outcomes of SLT1 and SLT2 yielded statistically similar results when both eyes were included. A future controlled prospective study would be best suited to address the shortcomings of this study. We believe the results from this retrospective analysis are sufficient to cast doubt on the assumption originating prior to the introduction of SLT that laser trabeculoplasty should only be repeated in eyes that show a successful response to initial treatment.

In conclusion, this is the first study to examine the efficacy of repeat SLT in eyes that exhibited a modest initial treatment response. For eyes with a successful initial response our results are similar to other reports. However, we have made the observation that even in eyes exhibiting a modest response to initial SLT, the proportion of those who responded successfully following repeat SLT was not significantly different from those who had a successful first response. This study highlights the need for a better understanding of SLT mechanisms of action and variables that may influence efficacy.