**5. The new trends of early detection of glaucoma and disease progression**

#### **5.1 Detecting of RNFL thinning and disease progression**

Evidence shows that RNFL abnormalities can often be evident without detectable VF damage [32–35]. Therefore, measuring the change of RNFL is likely to be useful in detecting early disease progression. Spectral-domain OCT is now the invaluable investigation tool for glaucoma patients because it can measure RNFL thickness reliably [36] with high sensitivity and specificity to detect glaucoma [37, 38] and its progression [39]. Commercially available software event-based algorithm, such as the Guided Progression Analysis (GPA, Carl Zeiss Meditec), can detect progressive RNFL thinning using RNFL thickness maps. A study has demonstrated that GPA can detect and visualise different patterns of progressive RNFL thinning [39]. Trend-Based Progression Analysis (TPA) is another algorithm for detecting progressive RNFL thinning by measuring the rate of change in RNFL thickness for each superpixel of the RNFL thickness map (50 x 50 superpixels). A study that involved 139 POAG patients (240 eyes) followed up for ≥5 years showed that progressive RNFL thinning determined by GPA and TPA was predictive of detectable functional decline in glaucoma. The study showed that TPA outperformed GPA in detecting more eyes with progressive RNFL thinning at a similar level of specificity (84.2% vs. 81.7% for TPA and GPA, respectively) [40]. Furthermore, TPA also provides visualisation of the distribution of the rate of RNFL thinning. It was suggested that the detection of progressive RNFL loss can serve as a biomarker to reflect disease deterioration behaviour and hence guide glaucoma management [40]. However, TPA is not without its limitations. A minimum of four follow-up visits is required for the construction of the TPA, and performance can be undermined with fewer visits. In situations where there are abrupt RNFL changes or in eyes with large test-retest variability, the event-based analysis may be more useful [41]. The authors concluded that TPA enhances but may not replace GPA for topographic analysis of RNFL thinning.

### **5.2 The dynamic target IOP, disease progression and quality of life: The LiGHT trial**

Clinical trials usually define a treatment IOP-lowering target. For instance, the OHT study aimed for an IOP lowering by 20% from baseline for patients in the treatment arm [3], whereas the Collaborative Normal Tension Glaucoma (CNTG) study targeted an IOP lowering by 30% from baseline [42]. The Laser in Glaucoma and Ocular Hypertension (LiGHT) trial is a multicentre RCT that compared eye drops versus selective laser trabeculoplasty as first-line treatment for POAG or OHT [43]. The study is unique with its well-constructed algorithm for detecting disease progression and guiding treatment escalation [44]. It has a novel approach to defining target IOP. Firstly, the target IOP is specific for each patient at baseline, based on disease severity and lifetime risk of loss of vision at recruitment (e.g. different target pressure and percentage IOP reduction according to the disease stratification suggested by Mills et al. [45]). Secondly, the IOP was adjustable based on IOP control and disease progression [44]. The disease progression (either glaucoma deterioration or conversion of OHT to POAG) was determined by a decision support software based on objective visual field and optic disc imaging criteria. Disease progression was defined as 'strong evidence', with the Humphrey GPA software showing 'likely progression' and/or HRT rim area > 1% per year (at *P* < 0.001), and 'less strong evidence' with GPA showing 'possible progression' and/or HRT rim area > 1% per year (at *P* < 0.01). 'Likely visual field progression' is the presence of three or more points on the GPA at <0.05 probability for change on three consecutive occasions, while 'possible visual field progression' is the same criterion but on only two consecutive occasions [44]. Optic disc progression was defined as the rate of neuroretinal rim loss exceeding 1% of baseline rim area/year on a minimum of five repeat HRT images, where this is equivalent to approximately twice the value of normal age-related rim area loss [22]. Following treatment escalation, there is a resetting of both the target IOP and visual field and optic disc baselines against which future assessments will be compared with.

Although the LiGHT trial is probably more complex in its target IOP setting algorithm compared with other glaucoma trials which defined treatment success based on the proportion of patients achieving a particular target percentage reduction of IOP, it resembled closer to our clinical practice. For instance, further IOP lowering beyond the 'target IOP' is probably required for patients with progressive disease. In some cases, patients might request to reduce medication use even when the target IOP is not achieved (e.g. OHT patients with IOP at 24 mmHg who do not want treatment and show no signs of POAG conversion). Furthermore, the LiGHT trial is also unique in that it included the evaluation of quality of life as an outcome measure. These are all novel features that might become important components for future glaucoma study design.
