Efficacy of Selective Laser Trabeculoplasty on Circadian Intraocular Pressure Following Trabeculectomy in Advanced Primary Open-angle Glaucoma

Précis: In about 50% of post-trabeculectomy (TE) eyes, selective laser trabeculoplasty (SLT) is effective in further lowering intraocular pressure (IOP). Purpose: To investigate the efficacy and safety of SLT in post-TE eyes, uncontrolled on maximum tolerated medication, and/or with progression of visual field loss. Patients and Methods: This retrospective study consecutively included post-TE eyes of patients diagnosed with primary open-angle glaucoma who had been treated with 360 degrees SLT and had a follow-up after 12 months. Primary endpoints were the reduction of mean diurnal intraocular pressure (mdIOP, mean of 6 measurements), peak IOP, and diurnal IOP fluctuations. Secondary outcomes were factors influencing IOP reduction, SLT success, and failure rates. Results: Forty-three eyes of 43 patients were included. During the first year, 10 eyes (23%) needed additional procedures to reduce mdIOP and were accounted as failures and excluded from final analysis. Of the remaining 33 eyes (77%) mdIOP [Q25, Q75] dropped from 15.2 [12.2 to 16.5] to 13.2 [11.6 to 15.3] mm Hg (P=0.027), 23 eyes (54%) showed a sufficient mdIOP reduction, 1 year after SLT. Conclusion: SLT is effective and safe in lowering mdIOP to target IOP in about 50% of eyes after prior incisional glaucoma surgery.

L owering intraocular pressure (IOP) to slow down or stop disease progression is still the primary goal of glaucoma therapy. 1 If maximum-tolerated medical therapy is not sufficient to control the disease, a laser procedure such as selective laser trabeculoplasty (SLT) 2,3 is often attempted before more invasive surgical interventions are performed. A number of studies have shown that SLT is effective and safe as the primary 4,5 as well as adjunctive therapy. 3,6 The delivery of electromagnetic energy to the trabecular meshwork (TM) is thought to induce inflammation, biochemical, and cellular changes resulting in improved trabecular aqueous outflow, which is the principal outflow pathway. Subsequently a decrease in IOP can be achieved. [7][8][9] The complete mode of action, however, is still not understood.
The current study investigated whether SLT is also effective and safe one year after treatment in patients with uncontrolled medically treated primary open-angle glaucoma (POAG) who had a prior trabeculectomy (post-TE). We hypothesized that SLT would be not effective at all in post-TE eyes because the TM and Schlemm canal had been bypassed by trabeculectomy for at least as long as filtration was functioning. As a result, TM function and outflow facility could probably not be further improved.

PATIENTS AND METHODS
A retrospective, electronic medical record review identified eyes of patients with POAG post-TE who consecutively had received SLT between May 2010 and October 2016 because mean diurnal IOP (mdIOP) became insufficiently controlled despite maximum tolerated IOP-lowering medication and/or showed progression of visual field loss. To be included patients had to be 18 years and older as well as having had a continuous follow-up of at least 12 months. The indication for SLT was decided at the discretion of the treating glaucoma specialist.
The study protocol was approved by the Institutional Review Board of the Medical Faculty Carl Gustav Carus of the Technische Universität Dresden, Germany and followed the tenets of the Declaration of Helsinki.
POAG was defined as showing damage to the inner layers of the retina on optical coherence tomography, a glaucomatous optic disc with diffuse or focal thinning of the neuroretinal rim resulting in characteristic glaucomatous cupping, and/or corresponding visual field defects, but no other ocular or systemic diseases that might cause visual field defects. All patients were of Caucasian ethnicity and admitted for a 1-day routine glaucoma work-up in a tertiary university hospital glaucoma service. For patients with 2 eyes eligible, the one with higher pre-SLT mdIOP was selected.
Baseline data were taken from the last 1-day routine glaucoma work-up before SLT and follow-up data from the one-day routine glaucoma work-up about 12 months after SLT. Recordings included age, sex, known duration of the disease, number of IOP lowering medication classes as well as previous surgeries. A comprehensive ophthalmic examination included refraction, best spectacle-corrected visual acuity (BCVA), slit-lamp biomicroscopy of the anterior segment, Goldmann applanation tonometry (GAT, average of 6 measurements, Haag-Streit, Koeniz, Switzerland), gonioscopy and dilated funduscopy with a 90-diopter lens. A complete glaucoma work-up included automated perimetry (Swedish interactive threshold algorithm standard 30-2 program; Carl Zeiss Meditec, Dublin, CA), confocal scanning laser ophthalmoscopy (HRT II, Heidelberg Engineering Inc., Heidelberg, Germany), scanning laser polarimetry (Nerve Fibre Analyzer GDxPRO, Carl Zeiss Meditec) and the optical coherence tomography glaucoma module (SPECTRALIS, Heidelberg Engineering Inc.).
Measurements of 24-hour diurnal IOP were taken at 1, 4, 7 and 10 PM in a sitting position at the slit-lamp with a Goldmann tonometer (Haag-Streit, Koeniz, Switzerland), at midnight in a supine position using the handheld Perkins MK3 tonometer (HS Clement Clark Ophthalmic, Haag-Streit UK) and at 7 AM again in a sitting position at the slitlamp before application of IOP-lowering medications. These are the usual measurement times for all glaucoma patients during a 1-day glaucoma work-up. Masked observers took one measurement at each time, as it is practice in daily clinical routine.
SLT was performed with the Ellex SLT laser (SoloTM, ellex Inc., Adelaide, Australia) using a Latina SLT goniolens (Ocular Instr., Bellevue, WA). The procedure has been described previously. 3,10,11 The initial energy level was set to 0.8 mJ and adapted according to the grade of angle pigmentation and bubble formation. About 100 nonoverlapping spots were applied in a single session to 360 degrees of the TM. Procedure details included the number of spots and total amount of energy used. No anti-inflammatory medications were applied postoperatively. The number of glaucoma medications and substances was not changed or discontinued after SLT.
Primary outcome measures were the reduction of medicated mdIOP (mean of 6 measurements including a measurement at midnight in a supine position), diurnal peak IOP and IOP fluctuations at 12 months. Diurnal IOP fluctuation was defined as the highest−(minus) the lowest IOP measured. Secondary outcomes were factors influencing IOP reduction, SLT success and failure rates, postoperative complications, and a possible deterioration in BCVA and field.
Success was defined as eyes that reached the individual target IOP without any change in glaucoma medication and with no need for glaucoma surgical interventions during the 1-year follow-up. In accordance with the Guidelines of the European Glaucoma Society, 12 target IOP was individually set as the upper limit of the IOP estimated to slow progression in such a way that vision-related quality of life in the expected lifetime of the patient, is maintained.
Failure was defined as eyes not reaching target IOP or showing progression and requiring a change in postoperative IOP-lowering medication or further surgical interventions during or at the 1-year follow-up, which was decided at the discretion of the treating glaucoma specialist. Data were excluded from final analysis at the time of further IOPlowering therapy or procedures.
Because patients included in this study were not untreated and a preoperative wash-out did not seem safe in such advanced cases, we did not define success or failure with an additional percentage of IOP reduction. The primary clinical goal was rather to reach target mdIOPs, reduce peak IOPs and IOP fluctuations.
Due to a non-normally distributed data set, nonparametric tests were used for analysis. The Wilcoxon test was applied for the comparison of preoperative and postoperative continuous data. The data were expressed as median and interquartile range [Q25 to Q75]. A multivariate linear regression analysis was performed to determine the influence of baseline factors [preoperative mdIOP/peak IOP/ IOP fluctuations, number of medications, pachymetry, and the severity of disease (MD, cup-disc ratio), age, sex, lens status, and the known duration of the disease] on SLT success. From this analysis, the main influencing factors were correlated with each investigated IOP value using Spearman correlation. A P-value <0.05 was considered as statistically significant.

Baseline Characteristics
A total of 43 eyes of 43 patients with POAG were included in this study. Demographic and baseline data are summarized in Table 1

Primary Outcome Measures
In the course of 12 months, 10 (23%) eyes (3 times increasing the application frequency of topical medication, 1 re-SLT, 6 surgical interventions) needed further steps to reduce mdIOP. These data were excluded from final analysis at the time of further IOP-lowering therapy or procedures and accounted as failures. Finally, 33 (77%) were evaluated at the 1-year follow-up. Medicated Fig. 1). Similarly, peak IOP was statistically significantly reduced, whereas diurnal IOP fluctuations were reduced but did not reach statistical significance ( Table 2, Mann-Whitney U test). After 1 year, 54% of eyes achieved success in sufficient IOP control. Another 10 eyes (23%) showed an increase in mdIOP and accounted as failures. The overall failure rate was 46%

Secondary Outcome Measures
In a linear regression analysis investigating influencing factors on SLT success, a higher preoperative mdIOP was not a factor predicting success in lowering mdIOP (β = −0.188; P = 0.435). A higher preoperative peak IOP (β = −0.944; P = 0.006) was the only factor predicting success in lowering peak IOP and higher preoperative IOP fluctuations (β = −0.874; P < 0.001) in lowering IOP fluctuations. The correlation analysis (Fig. 2) shows similar results (r = −0.237, P = 0.185, Spearman correlation) with no correlation between reduction of mdIOP and preoperative mdIOP. For the reduction of preoperative peak IOP, a weak correlation with a higher preoperative peak IOP was seen (r = −0.356; P = 0.042, Spearman correlation). For the reduction of diurnal IOP fluctuations, a moderate to strong correlation with higher pre-pachymetry, and the severity of disease, age, sex, lens status or the known duration of the disease did not have any influence on reducing mdIOP, peak IOP or diurnal IOP fluctuations.
No sustained postoperative complications, such as a rise in IOP, peripheral anterior synechiae or uveitis were seen. The most common adverse reaction was a mild anterior chamber inflammation, which returned to normal within 1 or 2 days. BCVA and MD stayed stable.

DISCUSSION
In this study, SLT was used as an additive to topical medical therapy to further reduce IOP in advanced post-TE eyes diagnosed with POAG in order to avoid or postpone subsequent incisional surgery or cyclodestructive procedures.
The main goal was to evaluate if SLT is still useful in post-TE eyes and whether the TM is functional enough, at least to some extent, to be activated by SLT. 8,9 Johnson et al 13 examined secondary changes in the TM and Schlemm canal after filtration surgery. They discovered that successful filtration surgery was associated with a decrease in the size of Schlemm canal due to underperfusion of the meshwork and therefore assumed that glaucoma could be more difficult to control if the filtration finally fails. The TM of post-TE eyes had not only been effectively bypassed by trabeculectomy for quite some time, but probably was already poorly functioning before trabeculectomy. Incisional glaucoma surgery is usually reserved for later stages of glaucoma to bypass an inadequately functioning outflow pathway. 14 Finally, post-TE eyes were again multi-treated eyes before SLT was performed.
Therefore, it could be assumed that SLT might have no effect in such eyes at all. However, the results of the current study allow the assumption of a still functional TM, at least   ALT indicates Argon laser trabeculoplasty; BCVA, best-corrected visual acuity; C/D ratio, cup to disc ratio; CAI, carbonic anhydrase inhibitors; MD, mean deviation; mdIOP, mean diurnal intraocular pressure; No., number; PGA, prostaglandin analogs; PSD, pattern standard deviation; SLT, selective laser trabeculoplasty; α-Ag, alpha-agonists; β-Bl, beta-blockers.
to some extent, which has the potential to be activated by SLT, even in post-TE eyes.
Zhang et al 15 examined the efficacy of SLT in post-TE eyes in a Chinese population. Similar to the current study, they treated post-TE POAG patients with 360 degrees SLT, who could not obtain target IOP through post-trabeculectomy medication. They defined SLT treatment success as > 20% IOP reduction compared with baseline and found a success rate of 77.7%. The reason for this outstandingly good result is probably the quite high preoperative IOP, which was on average 21.3 ± 3.4 mm Hg. They found a statistically significant decrease in IOP fluctuation from 4.1 ± 1.4 to 2.6 ± 1.1 mm Hg (P = 0.003) 9 months after SLT, whereas in the current study, though seeing a reduction in diurnal IOP fluctuations, the reduction did not reach statistical significance.
In a retrospective study, Sharpe et al 16 included 106 eyes and examined the efficacy of SLT following incisional glaucoma surgery (n = 53) versus a control group without prior incisional glaucoma surgery (n = 53). Contrary to the current study, patients were rather heterogeneous regarding to the type of open-angle glaucoma, type of incisional surgery as well as race, with about 64% being Afro-American patients. Another difference is that only 180 degrees inferiorly were treated with 50 adjacent laser spots and a retreatment of the superior 180 degrees was allowed. Success was defined as > 20% drop in IOP from pre-SLT baseline. Again, pre-SLT IOP was much higher than in the current study, 19.2 ± 4.3 vs. 15.2 ± 3.0 mm Hg in the prior incisional surgery group. About 28% (vs. 12% in our study) in the prior surgery group showed a 20% IOP reduction at the 1-year follow-up. As a conclusion of their investigation, they also found that SLT is efficacious after prior incisional glaucoma surgery. A higher pre-SLT IOP was associated with a larger IOP reduction, whereas in the current study, this correlation was not seen.
Although the current study found an mdIOP reduction of ≥ 20% in only 12% of patients receiving SLT post-TE, it has to be emphasized that the preoperative IOP was already quite low in most cases, which is sometimes still insufficient to control patients with advanced glaucoma. [17][18][19] Further reducing IOP in such cases is challenging. About half of the patients, 54% showed sufficient IOP reduction to reach their individual target IOP, with lower peak-IOPs and less IOPfluctuations one year after SLT. Any IOP reduction however, even of only 1 mm Hg, slows down progression as it was shown in the Early Manifest Glaucoma Trial. 17,20 The AGIS 18 showed, that if IOP continues to be <18 mm Hg at all visits, visual field progression is close to 0 over 6 years. Therefore, the reduction of daily peak IOP is extremely important. The AGIS 21 also showed that greater long-term IOP-fluctuations have been consistently associated with visual field progression. In the current study IOP fluctuations were reduced, the reduction did not quite reach statistical significance due to the rather low number of patients included, however. Considering the good safety profile and the rather fast and painless procedure, SLT can definitely be attempted in the population studied.
Several investigators 14,22,23 described a segmental variability of flow through the TM in glaucoma patients. Increasing alterations in the TM with disease progression do not uniformly affect the TM. Even in advanced disease, there might be some unaffected regions within the glaucomatous TM. The IOP reduction following SLT in post-TE eyes suggests at least some level of residual function in the TM that has the potential to be activated by SLT. Previous studies as well as the current study demonstrate that SLT efficacy is not correlated with the stage of disease, 3,9 which can also be explained by the segmental variability and some unaffected regions of trabecular outflow.
Interestingly, according to what was proven many times, 3,24 a higher preoperative mdIOP was not a predictor for success in the cohort studied as was shown with regression and correlation analysis. We hypothesize, that prostaglandins, which are common mediators in the inflammatory response caused by SLT might enhance uveoscleral outflow in such a way that the pressure gradient is not the driving force.
The strongest limitations of the study is the retrospective design. Decision for SLT as well as extra therapy was at the discretion of the treating glaucoma specialist. All procedures involved Caucasian patients and findings may not be applicable to other ethnicities. Another weakness is that we only had 1 day measurements of IOP at baseline and 1 year.
Strengths are the homogenous group regarding glaucoma diagnosis, ethnicity, sex, age, and preoperative mdIOP as well as the fact, that 6 diurnal IOPs, taken always at the same time of the day, including a measurement in the supine position were compared, rather than IOPs taken at nonstandardized points in time.
This allows assessing mdIOP, diurnal peak IOP and IOP fluctuations. Moreover, IOP-lowering medication was not changed during follow-up, allowing to assess the sole effect of SLT.
In conclusion, SLT could not only further reduce mdIOP in medicated post-TE eyes but could also reduce peak IOP and diurnal IOP fluctuations. The procedure is safe, fast, can be performed by comprehensive ophthalmologists and is able to achieve sufficient IOP control in about half of the cases.