Progression of visual field in patients with primary open‐angle glaucoma – ProgF study 1

Abstract

Purpose

To evaluate the visual field rate of progression of patients with treated ocular hypertension (OHT) and primary open‐angle glaucoma (POAG) in clinical practice, using the mean deviation (MD) and the visual field index (VFI).

Methods

Non‐interventional cohort study. From a large multicentre database representative of the French population, 441 eyes of 228 patients with treated OHT or POAG followed up at least 6 years with Humphrey 24.2 Sita‐Standard visual field examination at least twice a year were identified. From initial data, eyes were classified in five groups: 121 with OHT, 188 with early glaucoma (MD greater than −6 dB), 45 with moderate glaucoma (MD −6 to −12 dB), 41 with advanced glaucoma (MD −12 to −18 dB) and 46 with severe glaucoma (MD less than −18 dB). Rate of progression during the follow‐up period was calculated using the trend analysis of the Guided Progression Analysis software.

Results

The mean duration of follow‐up was 8.4 ± 2.7 years and the mean number of visual field, 18.4 ± 3.5. In eyes with OHT, rate of progression was −0.09 dB/year (−0.17%VFI/year). In eyes with POAG, rate of progression was −0.32 dB/year (−0.83%VFI/year) in eyes with early glaucoma, −0.52 dB/year (−1.81%VFI/year) in moderate glaucoma, −0.54 dB/year (−2.35%VFI/year) in advanced glaucoma and −0.45 dB/year (−1.97%VFI/year) in severe glaucoma. In eyes with POAG, a significant progression (p < 0.05) was detected in 159 of 320 eyes (49.7%) with trend analysis and 117 of 320 eyes (36.6%, likely progression) or 183 of 320 eyes (57.2%, possible and likely progression) with event analysis.

Conclusions

Primary open‐angle glaucoma is a progressive disease in the majority of patients despite cautioned treatment and follow‐up. The rate of progression varies greatly among subjects.

Introduction

Many large clinical trials have evaluated the time–course of visual field defects in glaucomatous subjects and demonstrated that most patients have visual field defects progression despite treatment and that the rates of progression varied greatly from one individual to other (Gliklich et al. 1989; Collaborative Normal‐Tension Glaucoma Study Group 1998; Membrey et al. 2000; Anderson et al. 2001; Eid et al. 2003; Leske et al. 2003; Heijl et al. 2009; Musch et al. 2009). Clinical trials evaluating the evolution in patients with treated glaucoma may differ from clinical practice for several reasons. Often, clinical trials include only a subset of subjects defined by strict inclusion criteria (e.g. early glaucoma naïve of any treatment, advanced glaucoma requiring filtering surgery). Patients included in the treated arm received a medical, laser or surgical treatment specified in advance and strictly codified, which may be different from what is done in clinical practice. The rhythm of follow‐up visits and the number of examinations performed to research progression at each follow‐up visit may also differ from what is done in practice. Finally, patient compliance to treatment and follow‐up examinations is often better for various reasons (essentially motivated patients agreed to participate in a clinical trial, patients could receive money for participating, doctors encourage greatly patients to comply with the prescribed treatment and with follow‐up visits, etc.).

Only a few numbers of studies have evaluated the progression of visual field defects in glaucoma patients treated and followed by usual care. Most of these studies were performed in the U.S., and one in Sweden (De Moraes et al. 2009, 2010, 2011, 2013a,b; Folgar et al. 2010; Prata et al. 2010; Heijl et al. 2013). Taking advantage of a large computerized database including the medical and ophthalmic records of hundreds of patients followed in two French centres for over 10 years, we aimed to study visual field progression and risk factors of progression in primary open‐angle glaucoma (POAG) and ocular hypertension (OHT) patients treated with the commonly used methods of treatment and followed in clinical ordinary practice. This should help the clinician to identify which patients require more care and might need a more aggressive treatment to achieve a better outcome. This first report focuses on the perimetric rate of progression at different stages of the disease.

Materials and Methods

Data source and population studied

All patients with glaucoma followed in the Glaucoma Unit of the Grenoble University Hospital and the Val de Grâce Military Hospital underwent a complete ophthalmic examination including visual field examinations as outpatients at least every 6 months. The patient's data have been recorded in a computer database since 1998. All patients provided both verbal and written informed consent to have their data recorded. Data collection was authorized by the Commission Nationale Informatique et Liberté (CNIL), the French information technology and personal data protection authority. Ethics committee approval to analyse the data was obtained from an institutional review board (IRB 5891 no. 2013‐25). Described research adhered to the tenets of the Declaration of Helsinki.

Each semiannual examination included at least a medical and ocular history, refraction, best‐corrected visual acuity, a Humphrey white on white visual field examination, Goldmann applanation tonometry, fundus examination and optic disc photographs. Patients are freely managed by the clinicians. In case of progression, clinicians were free to change medical treatment, perform argon or selective laser trabeculoplasty or perform incisional surgery. In case of cataract development, clinicians were free to perform cataract surgery.

By searching this database, we identified patients who fulfilled the following inclusion criteria: (i) diagnosis of POAG defined as IOP >21 mmHg before or during diurnal curve measurement with or without glaucoma medications, gonioscopically open angle, consecutive and reliable abnormal standard automated perimetry with abnormal results on the Glaucoma Hemifield Test and pattern standard deviation outside 95% of normal limits, and optic nerve damage (asymmetric cup/disc ratio >0.2, rim thinning, notching, excavation or retinal nerve fibre layer defect), or diagnosis of OHT defined as IOP >21 mmHg before or during diurnal curve measurement with or without glaucoma medications, gonioscopically open angle, two consecutive and reliable normal standard automated perimetry, and normal optic disc; (ii) at least 12 consecutive complete follow‐up visits spaced 6 ± 1 months; (iii) complete ophthalmic examinations including a Humphrey Visual Field Analyzer (Zeiss‐Humphrey Systems, Dublin, CA, USA) examination using the C‐24‐2 SITA standard strategy with reliable index during each of visits. A reliable visual field test was defined as having a <20% rate of fixation losses and <15% false‐positive and false‐negative results. Exclusion criteria were (i) other form of glaucoma, including normal tension glaucoma and pseudoexfoliative glaucoma; (ii) non‐glaucomatous ocular disease other than cataract; (iii) chronic non‐glaucomatous ocular medication; (iv) systemic diseases with possible effects on the visual field; and (v) cyclodestructive procedure previous or during the follow‐up period.

The following data were collected from the computerized database: name, sex, date of birth, ethnic origin, eye (right or left), glaucoma diagnosis classification, systemic medical condition history, ophthalmic condition history (eye, date, description), ophthalmic surgeries (eye, date, description), ophthalmic laser treatment (eye, date, description), systemic and ophthalmic medical treatment (name, indication, route, dose, frequency, eye if applicable, date of initiation, date of completion or ongoing), central corneal thickness, objective refraction, lens status (phakic, pseudophakic, aphakic), best‐corrected visual acuity, IOP, standard automated perimetry reliability indices, mean deviation (MD), pattern standard deviation, visual field index (VFI), trend and event analysis significance and trend analysis rate of progression. Data extraction was performed in March 2013.

Statistical analysis

The primary objective of this first report was to evaluate the visual field rate of progression of patients with treated OHT and POAG and to evaluate event and trend analysis capacity in detecting glaucoma progression at different stages. Beginning of the follow‐up period for a given patient was defined as the first visit that contributed records data from the patient. For statistical analysis, eyes were classified from initial data in five groups: OHT, early glaucoma (MD greater than −6 dB), moderate glaucoma (MD −6 to −12 dB), advanced glaucoma (MD −12 to −18 dB) and severe glaucoma (MD less than −18 dB). Rate of progression during the follow‐up period was calculated using the trend analysis of the Guided Progression Analysis software (Zeiss‐Humphrey Systems, Dublin, CA, USA). Event analysis was also performed with the Guided Progression Analysis software. Depending of the number of points showing significant progression and the significance of the progression, eyes are classified as ‘no progression’, ‘possible progression’ or ‘likely progression’ in the summary of the software event analysis. A significant decrease from baseline (two examinations) pattern deviation at three or more of the same test points on two consecutive tests is classified by the software as likely progression. A significant decrease from baseline (two examinations) pattern deviation at three or more of the same test points on three consecutive tests is classified by the software as likely progression.

Student's t‐test was used to compare means and percentages, and chi‐square tests were used for the analysis of dichotomous variables. The relationship between initial MD and rate of progression was evaluated using linear regression. Statistical significance was set at p ≤ 0.05. SPSS statistical software version 17.0 (SPSS Inc., Chicago, IL, USA) was used for analyses.

Results

Patient characteristics

In total, 441 eyes of 228 patients with OHT or POAG followed up at least 6 years were identified (121 with OHT, 188 with early glaucoma, 45 with moderate glaucoma, 41 with advanced glaucoma and 46 with severe glaucoma). The mean duration of follow‐up was 8.4 ± 2.7 years and the mean number of visual field, 18.4 ± 3.5. The mean follow‐up IOP were 16.4 ± 3.7, 15.1 ± 4.2, 15.0 ± 3.4, 14.1 ± 2.9 and 13.2 ± 3.1 mmHg, in eyes with OHT, early glaucoma, moderate glaucoma, advanced glaucoma and severe glaucoma, respectively. The subjects' demographic and ophthalmic data are shown in Table 1.

Table 1. Demographic and ophthalmic data

	Beginning of the follow‐up period	End of the follow‐up period
Demographic
Mean age (years)	65.9 ± 11.3	74.3 ± 11.3
Race (% Caucasian)	94.8	94.8
Sex (% Male)	59.1	59.1
Number of medications, n (%)
0	108 (24.5)	52 (11.8)
1	180 (40.8)	137 (31.1)
2	104 (23.6)	160 (36.3)
3	32 (7.3)	63 (14.3)
4	17 (3.8)	29 (6.6)
Lens status, n (%)
Phakic	342 (77.6)	238 (53.9)
Pseudophakic	97 (22.1)	201 (45.7)
Aphakic	2 (0.4)	2 (0.4)
Number of filtering surgeries, n (%)
0	308 (69.8)	257 (58.3)
1	120 (27.2)	161 (36.5)
2	12 (2.7)	22 (5.0)
3	1 (0.2)	1 (0.2)
Laser trabeculoplasties, n (%)
Yes	102 (23.1)	175 (39.7)
No	339 (76.9)	266 (60.3)

Visual field rate of progression

Ocular hypertension

At the beginning of the follow‐up period, mean MD and VFI were −1.38 ± 3.08 dB and 96.1 ± 8%. Rate of progression was −0.09 dB/year (−0.17%VFI/year) in eyes with OHT at the beginning of the study period. See Tables 2-4.

Table 2. Initial and final diagnosis: OHT, early glaucoma (MD greater than −6 dB), moderate glaucoma (MD −6 to −12 dB), advanced glaucoma (MD −12 to −18 dB) and severe glaucoma (MD less than −18 dB)

Diagnosis	Initial diagnosis (n)	Initial diagnosis (%)	Final diagnosis (n)	Final diagnosis (%)
1 (OHT)	121	27	75	17
2 (early)	188	43	184	42
3 (moderate)	45	10	64	15
4 (advanced)	41	9	45	10
5 (severe)	46	10	73	17

• MD = mean deviation; OHT = ocular hypertension.

Table 3. Rate of progression during the follow‐up period calculated using the trend analysis of the Guided Progression Analysis software

Rate of progression		Initial diagnosis
		OHT	Early glaucoma	Moderate glaucoma	Advanced glaucoma	Severe glaucoma
VFI (%/year)	Mean	−0.17	−0.83	−1.81	−2.35	−1.97
	SD	0.47	1.62	2.29	1.93	2.03
	95% CI	−0.25; −0.08	−1.06; −0.59	−2.49; −1.12	−2.95; −1.74	−2.57; −1.37
MD (dB/year)	Mean	−0.09	−0.32	−0.52	−0.54	−0.45
	SD	0.24	0.54	0.61	0.45	0.56
	95% CI	−0.13; −0.05	−0.39; −0.24	−0.70; −0.34	−0.68; −0.40	−0.61; −0.28

• OHT = ocular hypertension; SD = standard deviation; CI = confidence interval; VFI = visual field index; MD = mean deviation.

Table 4. Distribution of progression expressed as dB/year among patients

Rate of progression (dB/year)	Initial diagnosis and proportion of eyes, n (%)
	OHT	Early glaucoma	Moderate glaucoma	Advanced glaucoma	Severe glaucoma	All glaucoma
[+1, +∞[	0/121 (0)	0/188 (0)	0/45 (0)	0/41 (0)	0/46 (0)	0/320 (0)
[+0.5, +1[	0/121 (0)	1/188 (0.5)	1/45 (2.2)	0/41 (0)	0/46 (0)	2/320 (0.6)
[0, +0.5[	41/121 (33.9)	38/188 (20.2)	3/45 (6.6)	1/41 (2.4)	7/46 (15.2)	49/320 (15.3)
[−0.5, 0[	75/121 (62)	112/188 (59.6)	22/45 (48.9)	17/41 (41.5)	19/46 (41.3)	170/320 (53.1)
[−1, −0.5[	5/121 (4.1)	24/188 (12.7)	14/45 (31.1)	17/41 (41.5)	8/46 (17.4)	63/320 (19.7)
[−1.5, −1[	2/121 (1.7)	7/188 (3.7)	2/45 (4.4)	3/41 (7.3)	7/46 (15.2)	19/320 (5.9)
[−2, −1.5[	0/121 (0)	3/188 (1.6)	0/45 (0)	2/41 (4.9)	5/46 (10.9)	10/320 (3.1)
[−2.5, −2[	0/121 (0)	0/188 (0)	1/45 (2.2)	1/41 (2.4)	0/46 (0)	2/320 (0.6)
[−3, −2.5[	0/121 (0)	1/188 (0.5)	1/45 (2.2)	0/41 (0)	0/46 (0)	2/320 (0.6)
]−∞, −3[	0/121 (0)	2/188 (1)	1/45 (2.2)	0/41 (0)	0/46 (0)	3/320 (0.9)

• OHT = ocular hypertension.

Glaucoma

At the beginning of the follow‐up period, mean MD and VFI were −6.11 ± 7.7 dB and 83.5 ± 20.6%. Rate of progression was −0.32 dB/year (−0.783%VFI/year) in eyes with early glaucoma, −0.52 dB/year (−1.81%VFI/year) in moderate glaucoma, −0.54 dB/year (−2.35%VFI/year) in advanced glaucoma and −0.45 dB/year (−1.97%VFI/year) in severe glaucoma. See Tables 2-4. In all 320 patients with POAG at the beginning of the follow‐up period, the mean rate of progression was −0.40 ± 0.31 dB/year for the MD and −1.33 ± 1.89%/year for the VFI. We found a significant positive association between initial MD and rate of progression in early, moderate and advanced glaucoma (p < 0.04).

Event and trend analysis capacity to detect progression

In eyes with POAG, a significant progression (p < 0.05) was detected in 159 of 320 eyes (49.7%) with trend analysis and 117 of 320 eyes (36.6%, likely progression) or 183 of 320 eyes (57.2%, possible and likely progression) with event analysis (see Tables 5 and 6).

Table 5. Trend and event analysis significance during the follow‐up period

	OHT, n (%)	POAG, n (%)	All subjects, n (%)
Trend analysis (VFI ROP)
No significant slope	104 (85.9)	161 (50.3)	265 (60)
Significant slope	17 (14.1)	159 (49.7)	176 (40)
Event analysis
No progression	101 (83.5)	131 (40.9)	232 (53)
Possible progression	12 (9.9)	66 (20.6)	78 (18)
Likely progression	8 (6.6)	117 (36.5)	125 (28)
Not calculated (important)	0 (0)	6 (1.9)	6 (1)

• OHT = ocular hypertension; POAG = primary open‐angle glaucoma; VFI ROP = visual field index rate of progression.

Table 6. Trend and event analysis significance during the follow‐up period in patients with glaucoma

	Initial diagnosis
	Early glaucoma, n (%)	Moderate glaucoma, n (%)	Advanced glaucoma, n (%)	Severe glaucoma, n (%)
Trend analysis (VFI ROP)
No significant slope	98 (52.1)	26 (57.8)	17 (41.4)	20 (43.5)
Significant slope	90 (47.9)	19 (42.2)	24 (58.6)	26 (56.5)
Event analysis
No progression	71 (37.8)	19 (42.2)	14 (34.1)	27 (58.7)
Possible progression	40 (21.3)	12 (26.7)	6 (14.6)	8 (17.4)
Likely progression	77 (40.9)	12 (26.7)	17 (41.5)	9 (19.6)
Not calculated (important)	0 (0)	2 (4.4)	4 (8.9)	0 (0)

• VFI ROP = visual field index rate of progression.

Discussion

We evaluated the visual field rate of progression of a large cohort of French patients with OHT and POAG treated and followed in usual clinical care. During a mean period of 8.4 years, 49.7% of patients with glaucoma at the beginning of the follow‐up period showed significant progression with trend analysis, and 57.1% with event analysis. The mean rate of progression in patients with POAG during the whole follow‐up period was −0.40 ± 0.31 dB/year. The rate of progression varied considerably among patients who had nevertheless all the same type of glaucoma. A rate of progression >1 dB/year, which may be considered as a rather fast rate of progression, was found in 11.3% of the subjects with glaucoma. Patients with more advanced disease demonstrated a faster rate of progression compared to patients with early glaucoma.

This study has some strengths and weaknesses. Among the strengths, the number of subjects, the duration of the follow‐up period and the frequency of visual field tests are often higher than those of other clinical cohort studies conducted to date. Treatment and follow‐up of patients with glaucoma in France are often carried out in specialized hospitals, as private ophthalmologists are few (about 5500 for a population of 66 million, the optometry profession does not exist in France) and are primarily responsible for refractive tests, lenses prescription and cataract surgery (de Pouvourville & Chaine 2007). Therefore, we can estimate that the population involved in this study reflects an unselected population of patients with glaucoma rather than a subset of a specific population treated and followed in a tertiary glaucoma centre. In addition, no treatment algorithm was proposed or imposed to physicians. Finally, a large number of physicians specialized were involved in the treatment and monitoring of the patients included in this cohort: about 12 doctors for two centres. Some of these doctors are hospital and/or faculty doctors, others are private practitioners performing shifts in public hospitals. We can therefore estimate that this study reflects the ordinary care given to patients with glaucoma in France and that we have analysed the second reasonably representative population on this subject worldwide.

Only a few numbers of studies have evaluated the progression of visual field defects in glaucoma patients treated and followed by usual care (De Moraes et al. 2009, 2010, 2011, 2013a,b; Folgar et al. 2010; Prata et al. 2010; Heijl et al. 2013; Bertrand et al. 2014; Choi et al. 2014). One large cohort study was performed in the USA and has resulted in many analyses published, and one other was performed in Sweden. In a retrospective chart review of 583 patients with POAG and pseudo‐exfoliation glaucoma, Heijl et al. (2013) have measured a mean rate of progression of −0.80 dB/year over a mean follow‐up time of 7.8 years. Rate of progression also varied greatly among patients, 5.6% of patients progressing at rates worse than −2.5 dB/year. In multivariate analysis, pseudoexfoliation syndrome was not a significant risk factor of faster progression (slope −0.071; p = 0.21); however, we do not exactly know the rate of progression in POAG and pseudoexfoliation glaucoma subjects taken alone, preventing an accurate comparison to the results of our study. Apart from that, patient's characteristics were rather comparable to ours in terms of age and gender. In another retrospective chart review of 841 eye with different types of glaucoma followed over a mean period of 6.4 years, De Moraes et al. (2013a,b) have measured a mean rate of progression of −0.48 dB/year in POAG patients (275 eyes), faster in exfoliative glaucoma (−0.65 dB/year, 84 eyes) and slower in normal tension glaucoma (−0.33 dB/year, 81 eyes). Patient's characteristics were also comparable to ours in terms of age and gender, and most of their patients were of European ancestry (87%).

Prospective clinical trials have reported various rate of progression. By analysing the rate of progression of Early Manifest Glaucoma Trial patients randomized to the untreated control group over a 6 year period, Heijl et al. (2009) found a mean visual function loss of −1.08 dB/year overall, −1.31 dB/year in high‐tension glaucoma, −0.36 dB/year in normal tension glaucoma and −3.13 dB/year in exfoliative glaucoma. In the Canadian Glaucoma Study, the mean rate of progression was −0.54 dB/year in 45 patients having criteria of progression, and +0.06 dB/year in 153 patients do not having criteria of progression, leading to an estimated mean rate of progression of −0.08 dB/year (Chauhan et al. 2008, 2010). The present study has nevertheless several differences with the present study. The Canadian Glaucoma Study was an interventional study designed to evaluate the systemic risk factors associated with a higher risk of progression. Consequently, a very strict protocol for IOP control was applied to all patients included. Patients were followed up at 4‐month intervals, and patients reaching a priori defined criteria of progression underwent a 20% or greater additional reduction in IOP.

It should be noted that we have used a rather strict IOP criteria to define POAG (IOP >21 mmHg before or during diurnal curve measurement with or without glaucoma medications). This could have led to include more severe case of glaucoma, as previous studies have usually found a higher rate of progression in high‐tension glaucoma that in normal tension glaucoma (Heijl et al. 2009). Strictly, patients called ‘POAG’ in our study should be compared to patients called ‘high‐tension glaucoma’ in some of the previous studies conducted and above discussed (Heijl et al. 2009).

This study has several notable clinical implications. It shows that a certain proportion of glaucoma patients has a rapid rate of progression from the beginning of the disease. It also shows that patients with more advanced glaucoma have a faster rate of progression. We could therefore hypothesize that the rate of progression of the patients having a fast rate of progression at the beginning of the disease often remains unchanged throughout the course of the disease, despite a very regular follow‐up and care provided in a tertiary centre. Thus, those patients progress quickly to stages of moderate to advanced glaucoma, explaining that we found a faster rate of progression in patients with advanced glaucoma. Those patients are at risk to have a significant impact on their visual function and quality of life, and eventually to become blind. For these patients, with a fast rate of progression from the beginning of the disease, algorithms of treatment commonly used and advocating a gradual escalation of treatment are probably partially inadequate (Terminology and Guidelines for Glaucoma 2008; American Academy of Ophthalmology Glaucoma Panel 2010). It would probably be better to consider a more aggressive treatment early in the course of the disease, or at least as soon as the fast rate of progression is measured. In contrast, another significant part of glaucomatous subjects has a slow or nearly zero rate of progression. In these patients, less frequent follow‐up could probably be enough, and we could be less prompt to consider a therapeutic climbing.

In conclusion, this study shows that glaucoma is a progressive disease in the majority of patients despite cautioned treatment and follow‐up. The rate of progression varies greatly from one subject to another. Therefore, the initial MD and MD course observed at the beginning of the disease determine the future evolution profile, the risk of visual impairment and progression to blindness. Rather than applying a treatment and follow‐up algorithm uniform in all patients, it could therefore be pertinent to consider the rate of progression measured in the early years of the follow‐up, with consequent aggressive treatment immediately and a very close follow‐up for those progressing quickly and vice versa for those progressing slowly.