figure a

Introduction

The improvement of health care over the past decades has increased the life span of Asians, which might have contributed to the higher incidence of age-related diseases such as glaucoma. Asians account for 59.76% of the total world population [1]. Glaucoma is known as the most common cause of blindness in Asia [2, 3]. Previous studies estimated that the higher prevalence of blindness due to glaucoma in Asia is owing to the higher incidence of primary angle-closure glaucoma (PACG) [4, 5].

Asia consists of a heterogenous population, making it a melting pot [6]. Accordingly, the prevalence of glaucoma varies among Asian populations. Malays are the people who reside in the Malay Archipelago, which includes Malaysia, Indonesia, Singapore, Thailand, Brunei, and the Philippines. They compose 4.2% of the world population [7]. Although approximately 300 million people of Malay ethnicity live in Asia [8], information on the burden, causes, risk factors, and epidemiology of blinding eye diseases in this ethnic group is lacking. Most knowledge about eye diseases has been derived from Chinese, Japanese, and Indian populations [6, 9,10,11], but little knowledge is known in the Malay population. Thus, so far, the Singapore Malay Eye Study provides the only available epidemiological data on ocular diseases in Malays [12]. The prevalence of glaucoma was reported to be 4.6, with prevalence rates of 3.2 and 0.2 for primary open-angle glaucoma (POAG) and PACG, respectively [12]. Thus, understanding glaucoma in this subethnic group of Asians is of profound importance.

On the basis of the National Eye Survey II conducted in Malaysia to determine the causes of blindness and visual impairment in persons aged ≥ 50 years, glaucoma is responsible for 6.6% of all cases of blindness [13]. While most other causes of blindness in this survey are reversible blindness, 58% were due to cataract, 5% were due to cataract surgery-related complications, and 3.5% were due to corneal opacity [13]. As the aging population is increasing exponentially, understanding the characteristic of the progression rate and factors affecting visual field progression in Malay patients with glaucoma is crucial. According to Department of Statistics Malaysia, Malays account for 69.8% of the population in Malaysia [14]. To prepare Malaysia to be an aging country, these data are crucial for rehabilitation and prevention of blindness among older adults. The main objective of this study was to determine the 5-year survival rate and factors affecting visual field progression in Malay patients with POAG and PACG.

Materials and methods

A retrospective cohort record review was conducted in patients with POAG and PACG who were treated and followed up at Hospital Universiti Sains Malaysia (HUSM), a tertiary center in Kelantan, in the east coast of Malaysia. Kelantan is one of the states whose populations are mostly Malays [15]. This study received ethical approval from the research and ethics committee of the School of Medical Sciences, Universiti Sains Malaysia (reference code, USM/JePEM/16090340).

Potential patients were identified from the Malay Glaucoma Eye Study (MaGES) database created in 2014. This database was established to study the modifiable risk factors of POAG and PACG in Malay patients. POAG is characterized as a chronic, slowly progressive visual field loss and optic nerve cupping, often associated with an elevated intraocular pressure (IOP) and visually open anterior chamber angles on gonioscopy, without any underlying secondary ocular disease [16]. PACG is defined as an eye condition that presents with 180º or more occludable drainage angle and features that indicate trabecular obstruction by the peripheral iris, such as an increased IOP of > 21 mmHg, peripheral anterior synechiae, iris whirling, “glaucomflecken” lens opacities, or excessive pigment deposition on the trabecular surface with evidence of glaucomatous optic neuropathy [17]. In our database, the Malay patients were Muslims who descended from at least three generations, without interracial marriages, spoke Malay as their first language, and practiced Malay customs. They had already undergone screening for eligibility, which included the use of a pedigree chart to exclude any potential interracial marriages in three generations of their lineage, before entry into the database.

The MaGES database is comprised of 750 Malay patients who were recruited from four tertiary centers in Malaysia, namely, HUSM, Hospital Sultanah Bahiyah, Hospital Sultanah Nur Zahirah, and Hospital Kuala Lumpur. Of the 750 patients, only 265 were from HUSM. MaGES database only included Malay patients who fulfilled the diagnosis of POAG and PACG with good follow-up record and achieved target pressure for the last one year. The recruited patients must provide two reliable and reproducible visual fields within 3 months of the recruitment period. Patients with underlying retinal, media opacities, neuro-ophthalmology, or other systemic neurological disease that interferes with visual field interpretation were excluded. Myopic patients with refractive error of ≥ 4 diopters were also excluded. Those with dementia, chronic or persistent disorder of the mental processes, brain disease or injury, memory disorder, and psychotic instability were also excluded.

The medical records of the selected patients were traced. All the recruited patients must have at least six reliable and reproducible visual fields using the Humphrey visual field (HVF) based on the Swedish Interactive Threshold Algorithm standard 24–2 analysis. The reliability of the HVF is based on the reliability index of the visual field. A reliable visual field includes a fixation loss of < 20%, false-negative result of < 33%, and false-positive result of < 33% [18]. Of the 265 patients identified from the database, 28 were diagnosed after December 31, 2015, six failed to complete 5 years of follow-up or a minimum of 18 eye clinic visits, and seven failed to produce six reliable and reproducible HVFs. A total of 222 patients (222 eyes) who had primary glaucoma (POAG and PACG) and who were diagnosed and followed up for a minimum period of 5 years (minimum of 18 eye clinic visits) between January 1, 2009, and December 31, 2014, were included. They were also required to be followed up for another year between January 1, 2015, and December 31, 2015.

Scoring of the HVF was performed by two glaucoma consultants (LS and AY) using the Hodapp–Parrish–Anderson (HPA) scoring system [19] and Advanced Glaucoma Intervention Study (AGIS) criteria [20]. They were blinded from each other’s scoring. Visual field progression is defined when the HPA score indicates changes in severity [21] and the worsening of 4 unit based on AGIS score [22, 23]. Time of progression refers to the duration between the time of the initial diagnosis and the time when the first VF progression was detected (in years). Only the right eye was selected for HVF scoring if both eyes were eligible for recruitment. Based on the definition of progression, the patients were divided into two groups: a progression group and a non-progression group.

The medical records of the selected patients were traced. Data were collected by three investigators (WEA, HAS, and DTSJ) and divided into ocular and systemic parameters. Ocular parameters included the presence of ocular pain, red eye, and symptoms of acute angle-closure (AAC); visual acuity and IOP at diagnosis; laterality; vertical cup-to-disc ratio (VCDR); disc hemorrhage; central corneal thickness; and treatment given. The IOPs at diagnosis (baseline); 12, 24, and 36 months after diagnosis; and the time of recruitment were also recorded. The presence of other systemic comorbidities such as hypertension, diabetes mellitus, and hyperlipidemia were noted. These parameters were used as potential prognostic factors in the analysis. Any patients with missing data of > 30% were excluded. None of our patients were excluded owing to this reason.

All data were entered into the Statistical Program for Social Science (SPSS) version 22.0 software. They were checked and cleaned to ensure accurate documentation and to eliminate any missing or erroneous values (NMY and NSB). The SPSS and Statistical Data Analysis (STATA) version 22.0 software were used for the statistical analysis. For all the numerical variables, the normality distribution was assessed. A Kaplan–Meier survival probability curve was plotted to estimate the progression of glaucoma on the basis of the visual field progression [24]. Multivariable analysis using the Cox proportional hazard regression model was used to identify the prognostic factor for the progression of glaucoma. All covariates with p values < 0.25 in the univariable analysis or that were clinically important were included in the multivariable analysis. The parsimonious model refers to the simplest model with the fewest possible number of significant variables.

Results

In this study, 110 patients with POAG and 112 patients with PACG were recruited. The study included 113 (50.9%) female and 109 (49.1%) male patients who were followed up for a mean (SD) duration of 6.5 (3.6) years. Patients with visual field progression were significantly older at the time of recruitment and followed up for a shorter duration (Table 1). Most patients had a bilateral disease. A significantly higher percentage of those with a history suggestive of acute IOP elevation developed visual field progression (Table 2). Disc hemorrhage was found in 35 eyes, of which 54.3% developed visual field progression (Table 2).

Table 1 Comparison of demographic characteristics of Malay patients with and without visual field progression
Full size table
Table 2 Comparison of ocular parameters among Malay patients with and without visual field progression
Full size table

Based on two event analyses of HVF, 63 eyes (28.4%) were confirmed to have visual field progression. The Kaplan–Meier analysis estimated that the cumulative probability of non-progression was 79.8% at 5 years of follow-up (Fig. 1). The estimated median time of progression was 12.0 (95% confidence interval [CI], 8.5, 15.5) years, which indicated 50% of the study population were expected to progress at this timeline. No significant difference in cumulative probability of non-progression was found between the patients with POAG (67.4%) and those with PACG (70.2%) (Fig. 2, Table 3). In general, PACG progressed slower than POAG but not significantly (Tables 2 and 3). However, the female patients had a lower cumulative probability of non-progression (67.5%) than the male patients (70.1%) at 5 years, but the difference was not statistically significant (Tables 1 and 3).

Fig. 1
figure 1

Kaplan–Meier survival curve for visual field progression in Malay patients with primary glaucoma

Full size image
Fig. 2
figure 2

Comparison of cumulative probability of non-progression between Malay patients with POAG and PACG

Full size image
Table 3 Five-year survival probability and estimated time for visual field progression in Malay patients with POAG and PACG
Full size table

The Malay patients with a history of ocular pain and red eyes had a significantly higher probability of progression based on the Kaplan–Meier survival analysis (Tables 2 and 3). The probability of progression was also higher in the patients with disc hemorrhage (Tables 2 and 3). These predictors were significant in the simple Cox regression analysis, but only disc hemorrhage and ocular pain remained significant in the multiple Cox regression analysis (Table 4). The presence of disc hemorrhage increased the risk of visual field progression by 2.8-folds (95% CI, 1.61–4.76). The patients with a history of ocular pain had a 2.5-folds (95% CI, 1.37–4.44) increased risk of visual progression.

Table 4 Prognostic factors for visual field progression in Malay patients with POAG and PACG
Full size table

Discussion

The progression of glaucoma is subtle and often asymptomatic, which leads to late detection [4]. The absence of a standardized definition of progression further complicates the issue. Visual field changes are the features most commonly adopted in the definition of glaucoma progression [25]. Currently, with the popularity of optical coherent tomography (OCT), progression can now be detected on the basis of structural changes [26].

Visual field progression can be defined using an event- or trend-based analysis. In the present study, progression was defined based on the agreement of the results of two event-based analysis, namely, the AGIS score and HPA classification. A total of 63 patients (28.4%) with primary glaucoma were found to show visual field progression after 6 years of follow-up. The Kaplan–Meier analysis revealed that the cumulative probability of survival of the patients with no progression was 79.8% at 5 years, which was interpreted as an estimation of visual field progression of 20.2% at 5 years of follow-up in the patients with POAG and PACG. So far, no similar study has combined these two most common types of primary glaucoma.

Survival analysis for visual field progression in patients with POAG is often performed as the outcome of treatment or surgical intervention in many large prospective randomized controlled trials such the AGIS, Early Manifest Glaucoma Treatment Study (EMGTS), and Collaborative Initial Glaucoma Treatment Study [20, 27, 28]. Chen et al. found that 14.6% of patients with OAG became blind after 15 years of follow-up on the basis of their Kaplan–Meier survival analysis results [2]. Hattenhauer et al. found that 27% of patients with POAG became blind after 20 years of follow-up [29]. By using a special software analysis, many studies defined progression based on the rate of progression [30,31,32]. Thus, direct comparison between studies is difficult because they used different definitions of progression.

Visual field defect in PACG eyes is more diffuse [33]. Both POAG and PACG eyes showed more pronounced damage in superior hemifield but with higher tendency in POAG eyes [33, 34]. Based on mean deviation, the rate of visual field progression is faster in Asian patients with POAG [34,35,36]. The rate of progression predicts visual field defect per year and differs from estimation of cumulative probability of progression in survival analysis. However, if one presumed that the rate of progression is constant, the summation over the years may represent the cumulative probability of progression. Perhaps the cumulative effect of the rate of progression causes shorter median time to progress in Malay patients with POAG. Baseline visual field defect determines the subsequent progression. POAG eyes showed faster rate of progression compared to PACG eyes with similar baseline visual field defect [35]. However, baseline visual field of both type of glaucoma was not matched in the present study.

In Asian patients, PACG is believed to progress faster than in Caucasian patients [37, 384]. PACG is believed to be more aggressive in Malays than in Chinese [39, 40]. However, no significant difference in the percentage of progression was found between POAG (31.8%) and PACG (25.0%) in the present study. In fact, patients with PACG showed a higher cumulative probability of non-progression at 5 years of follow-up, with a longer estimated mean time of progression than POAG but with no statistically significant difference. Quek et al. reported that 32.5% of Chinese patients with PACG developed visual field progression after 10 years of follow-up [41]. On the basis of an indirect comparison, our study showed a slightly higher cumulative probability of non-progression at 5 and 10 years than their study. However, this analysis was not described in detail in their study. Apart from this discrepancy, the differences in follow-up duration and definition of progression make direct comparison impossible. Different methods of assessment cause a discrepancy in the detection of visual field progression [23]. Our study adopted a stricter definition of progression, which might have contributed to the lower percentage of visual field progression.

In general, PACG is more common in women, and men are more predisposed to POAG [42, 43]. However, no evidence has been found to support the role of sex in the progression and severity of glaucoma. Female patients were more likely to develop visual progression in the present study. Though not statistically significant, they had a lower cumulative probability of non-progression (67.5% at 5 years) but longer estimated time (15.1 years) of progression than men. Women have longer life spans, which increase the likelihood of progression at a later age [44,45,46].

In this study, disc hemorrhage was a strong predictor of progression in Malay patients with primary glaucoma. The cumulative probability of non-progression was 40.5% at 5 years in the patients who presented with disc hemorrhage. Drance et al. found that the presence of disc hemorrhage reduced the survival time to 1187 days compared with the 2159 days in those without disc hemorrhage [47]. Disc hemorrhage was also found to increase the risk of progression by 2% in the EMGTS [44]. It is known to cause a localized progression of visual field defect [45, 46]. In the present study, disc hemorrhage increased the risk of progression by 2.8-folds (95%CI 1.61, 4.76), which is almost similar to the finding reported by Drance et al. [47]. Disc hemorrhage was strongly associated with glaucoma progression in many prospective cohort studies [48, 51,52,53].

POAG is almost asymptomatic; however, many cases of sudden elevation of IOP have been reported [54]. Elevation of IOP may present with painful red eye [55]. PACG is more symptomatic especially in patients with a history of AAC [56]. However, not all patients with AAC will develop PACG if proper prompt treatment is given [57]. On the other hand, many cases of asymptomatic PACG that behave similarly to POAG have been reported especially in Asians [38, 39, 58]. Such cases were also observed in this study, with only a third of the patients with PACG presenting with a history of AAC. Fluctuation of IOP is known to cause further glaucomatous damage [59, 60]. In this study, a significant difference in 5-year progression of visual field was found between the patients with and those without eye pain.

The risk of progression was 2.5-folds (95% CI, 1.37–4.44) higher in the presence of eye pain but with no statistical significance. The presence of red eye increased the risk of progression of PACG by 2.7-folds (95% CI, 1.07–6.93) in a retrospective study involving all spectrums of primary angle-closure [39]. Including patients with POAG probably reduced the significant role of red eye in the present study. On the other hand, the incidence of eye pain and red eye may overlap with symptoms of AAC in patients with PACG. The retrospective nature of this study might likely be responsible for this bias.

In addition, owing to the nature of this study, several confounding factors could not be included because of missing data. This includes IOP fluctuation during follow-up. As structural changes preceded the visual field changes, the changes may be inaccurate because of the discrepancy during clinical observation. Our retrospective review was conducted during the time when diagnostic imaging was not available for the optic nerve head. A prospective cohort study will provide a better understanding of progression in Malay patients. PACG has different presentations from POAG; thus, a separate analysis of the prognostic factors for progression in patients with PACG is important.

In conclusion, the prevalence rate of visual field progression in Malay patients with primary glaucoma was 28.4%, and the median time for progression was 12.0 years (95% CI, 8.5–15.5). The visual field progression in Malay patients was almost similar with that in other Asian populations. The risk of progression was higher in the eyes with disc hemorrhage and history of eye pain, which warrant early detection and aggressive management.