Visual Field Outcomes of Two Major Glaucoma Trials

Visual Field Outcomes From the Multicenter, Randomized Controlled Laser in Glaucoma and Ocular Hypertension Trial (LiGHT)

Wright DM, Konstantakopoulou E, Montesano G, et al; Laser in Glaucoma and Ocular Hypertension Trial (LiGHT) Study Group¹
Industry support: D.P.C., financial support (Allergan, Bayer, Santen, Thea)

ABSTRACT SUMMARY

The prospective, multicenter, randomized controlled Laser in Glaucoma and Ocular Hypertension (LIGHT) trial analyzed the health-related quality of life and clinical effectiveness of initial treatment with selective laser trabeculoplasty (SLT) compared to IOP-lowering drops over 6 years in 718 patients newly diagnosed with and previously untreated for ocular hypertension or open-angle glaucoma.² Participants were randomly assigned to either the medical therapy–first group or the SLT-first group. Subsequent treatment changes were determined by a computerized algorithm. At each visit, individuals underwent Humphrey visual field (HVF; Carl Zeiss Meditec) testing with the 24-2 Swedish interactive thresholding algorithm standard.

The trial¹ included 11,563 VFs from 1,178 eyes of 688 patients after unreliable fields. Patients who had undergone cataract or trabeculectomy surgery were excluded. VF progression rates for both total deviation (TD) and pattern deviation (PD), recalculated from all 52 points, were estimated using hierarchical linear models with a trend-based analysis.

Overall, the mean TD and PD progression rates did not differ statistically between the medication (-0.25 dB/y) and SLT (-0.19 dB/y) groups. A greater proportion of eyes in the medication group, however, showed rapid TD progression, defined as moderate (< -0.5 dB/y) and fast (< -1 dB/y) progression (risk ratio, 1.43 [95% CI, 1.11–1.83]; P < .005). PD progression was not statistically significant in either group. The eyes experiencing moderate or rapid progression had lower average sensitivity at baseline and lower set IOP targets.

DISCUSSION

What are the study’s potential implications for real-world practice?

The VF outcome analysis averaged individual pointwise rates of VF progression to produce global rate estimates (TD and PD), which the authors argue retain more information than relying solely on global HVF measures such as mean deviation or pattern standard deviation. Separating progression rates into slow, moderate, and fast revealed that differences between SLT and medical therapy were evident only in patients experiencing moderate or fast progression who had identical IOP targets. Because clinical trial patients tend to adhere to prescribed medical therapy, the difference may be greater in the real world.

Why did the SLT-first group experience slower progression?

One possibility is that medically treated IOP naturally has peak and trough effects, leading to variability that may not be captured by typical IOP measurements in the clinic.³ Likewise, IOP control with medications relies on treatment adherence. Among LIGHT participants, self-reported adherence ranged from 75% to 99%.² Actual dose monitoring was not performed, however, and previous studies of real-world adherence to medications have shown lower adherence rates.^4,5 SLT may also achieve better 24-hour IOP control, reducing IOP variability.⁶ LIGHT¹ supports the concept that, with respect to preventing glaucomatous progression, an IOP of 15 mm Hg after a procedure without medications is not equivalent to an IOP of 15 mm Hg controlled with multiple medications without a procedure.

Five-Year Visual Field Outcomes of the HORIZON Trial

Montesano G, Ometto G, Ahmed IIK, et al⁷
Industry support: Study financially supported by Ivantis/Alcon; G.M., G.O., P.Y.R., D.F.C., I.K.A., and G.G., Consultants (Alcon)

ABSTRACT SUMMARY

The multicenter HORIZON randomized controlled trial compared 5-year results among patients with mild to moderate primary open-angle glaucoma (POAG) who underwent cataract surgery alone (CSA) or in combination with the placement of a Hydrus Microstent (CSH; Alcon). The trial found a sustained, significant reduction in the need for secondary glaucoma surgery and medication use in the CSH patients.⁸

The post hoc analysis⁷ compared the rate of progression (ROP) of VF damage in 3,580 reliable fields from 187 CSA patients and 369 CSH patients. HVF testing with the 24-2 Swedish interactive thresholding algorithm standard was performed annually after surgery. A linear mixed-effects model similar to that used in LIGHT was used to calculate the point-wise sensitivity over time from the surgery date. The ROP was reported by eye, individual location, and regions in the VF (Garway-Heath clusters⁹). A time to VF progression event was analyzed with predefined event thresholds of 2.5-, 3.5-, 4.5-, and 5.5-dB changes from baseline.

The overall ROP of VF damage was significantly slower in the CSH group (-0.26 dB/y) than in the CSA group (-0.49 dB/y). All clusters except the macular and inferior paracentral ones had a significantly slower ROP in the CSH group. The CSH arm also had significantly slower ROPs for the fastest progressing location (-1.55 vs 2.48 dB/y) and cluster (-0.79 vs -1.37 dB/y). Montesano et al calculated a time-weighted average daytime IOP based on clinic measurements that accounted for the frequency and spacing of IOP measurements. The daytime IOP was slightly lower in the CSH versus CSA group (16.37 vs 17.22 mm Hg); modeling showed the difference explained only 17% of the effect of the device on VF preservation.

The difference in time to VF progression was significant only for the progression threshold of -5.5 dB, indicating the presence of patients with similarly slow and moderate progression in both groups but a reduction in fast progression in the CSH group.

The findings indicate that the Hydrus Microstent can reduce the ROP in POAG patients, especially those experiencing fast VF progression.

DISCUSSION

Which findings on VF progression are clinically meaningful?

Montesano et al argue that a -0.26 dB/y ROP reduction after cataract surgery and Hydrus implantation is clinically meaningful because -0.5 dB/y is a commonly reported rate of deterioration in glaucoma cohorts.¹⁰ The difference between HORIZON treatment arms appears to be greater than the difference between the SLT and medication groups in LIGHT (-0.19 vs -0.25 dB/y),¹ potentially owing to the strict standardized treatment regimen in LIGHT and differences in the disease severity mix in an SLT-first cohort.

The time-to-event analysis in the HORIZON trial showed a significant difference between treatment arms in the time to reach 5.5 dB of VF progression. This suggests the implant can reduce the risk of extremely fast VF progression and is consistent with LIGHT, in which the medication-dependent arm showed higher rates of fast to moderate VF progression in the presence of equal in-office IOP measurements.¹ Another implication of the HORIZON trial is that the observed difference in the ROP between groups is partially driven by the fastest-progressing eyes.

Are there additional implications for real-world practice?

Unlike in LIGHT, HORIZON patients were treated according to standard clinical practice; their nonmasked treating physicians were free to escalate or remove medications or to perform additional laser treatment or surgery as they deemed necessary. The HORIZON results could therefore be considered more representative of real-world practice and translatable to clinic settings than those of LIGHT. There could also have been some treatment bias between the two arms of the HORIZON study. Montesano et al noted that the Hydrus arm had a lower average daytime IOP and that this difference of approximately 1 mm Hg in average daytime IOP explained only 17% of the reduction in VF deterioration, an effect that existed despite the lower number of medications used in the CSH group. As with LIGHT, these results may suggest that VF protection resulting from stent placement does not arise solely from a reduction of daytime IOP and that perhaps IOP stabilized by an intervention that is not affected by patient adherence (eg, dosing and peak/trough effects) is important for slowing long-term VF progression.

HORIZON participants included individuals with early or moderate POAG who were scheduled for cataract surgery and already receiving medical treatment (1–4 topical drops). LIGHT participants, in contrast, were newly diagnosed with POAG or ocular hypertension and naïve to treatment. It may not be possible to generalize the LIGHT or HORIZON results to patients with other forms of glaucoma or more advanced disease. Thus far, the 2-year VF outcomes in the Treatment of Advanced Glaucoma Study (TAGS) comparing trabeculectomy to medication for the treatment of newly diagnosed advanced glaucoma have not reached statistical significance.¹¹ Nonetheless, VF outcomes from LIGHT and HORIZON provide compelling evidence of the benefits of procedural IOP lowering compared to medication alone.