Beruflich Dokumente
Kultur Dokumente
Summary
Background Primary angle-closure glaucoma is a leading cause of irreversible blindness worldwide. In early-stage Lancet 2016; 388: 1389–97
disease, intraocular pressure is raised without visual loss. Because the crystalline lens has a major mechanistic role, See Editorial page 1349
lens extraction might be a useful initial treatment. See Comment page 1352
Centre for Public Health,
Methods From Jan 8, 2009, to Dec 28, 2011, we enrolled patients from 30 hospital eye services in five countries. Queen’s University Belfast,
Belfast, UK
Randomisation was done by a web-based application. Patients were assigned to undergo clear-lens extraction or
(Prof A Azuara-Blanco PhD);
receive standard care with laser peripheral iridotomy and topical medical treatment. Eligible patients were aged School of Medicine, University
50 years or older, did not have cataracts, and had newly diagnosed primary angle closure with intraocular pressure of St Andrews, St Andrews, UK
30 mm Hg or greater or primary angle-closure glaucoma. The co-primary endpoints were patient-reported health (J Burr MD); Health Services
Research Unit
status, intraocular pressure, and incremental cost-eff ectiveness ratio per quality-adjusted life-year gained 36 months
(Prof C Ramsay PhD,
after treatment. Analysis was by intention to treat. This study is registered, number ISRCTN44464607. D Cooper PhD, G Scotland PhD,
C Cochrane MSc,
Findings Of 419 participants enrolled, 155 had primary angle closure and 263 primary angle-closure glaucoma. Prof J Norrie PhD), Health
Economics Research Unit
208 were assigned to clear-lens extraction and 211 to standard care, of whom 351 (84%) had complete data on health
(G Scotland, M Javanbakht PhD),
status and 366 (87%) on intraocular pressure. The mean health status score (0·87 [SD 0·12]), assessed with the and Centre for Health Care
European Quality of Life-5 Dimensions questionnaire, was 0·052 higher (95% CI 0·015–0·088, p=0·005) and mean RandomisedTrials
intraocular pressure (16·6 [SD 3·5] mm Hg) 1·18 mm Hg lower (95% CI –1·99 to –0·38, p=0⋅004) after clear-lens (Prof J Norrie), University of
Aberdeen, Aberdeen, UK; NIHR
extraction than after standard care. The incremental cost-eff ectiveness ratio was £14 284 for initial lens extraction Biomedical Research Centre,
versus standard care. Irreversible loss of vision occurred in one participant who underwent clear-lens extraction and Moorfields Eye Hospital and
three who received standard care. No patients had serious adverse events. University College London, UK
(Prof P J Foster PhD); and
Johns HopkinsWilmer Eye
Interpretation Clear-lens extraction showed greater effi cacy and was more cost-effective than laser peripheral Institute, Baltimore, MD, USA
iridotomy, and should be considered as an option for first-line treatment. (Prof D S Friedman PhD)
Correspondence to:
Funding Medical Research Council. Prof Augusto Azuara-Blanco,
Institute of Clinical Sciences,
Block B, Grosvenor Road,
Copyright © The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY license. Belfast BT12 6BA, UK
a.azuara-blanco@qub.ac.uk
Introduction In the early stage of the disease, primary angle closure
WHO ranks glaucoma as the leading cause of irreversible is accompanied by high intraocular pressure but no
blindness,1 and prevalence is expected to increase visual loss. The standard of care for primary angle closure
substantially: compared with 20 million people who have and primary angle-closure glaucoma is laser peripheral
primary angle-closure glaucoma now, by 2040, 34 million iridotomy to open the drainage pathways and medical
people will be affected, of whom 5⋅3 million will be blind.2 management with eye drops to reduce intraocular
The prevalence of primary angle-closure glaucoma is pressure.7 If the disease remains uncontrolled, surgery,
highest in people of east Asian origin.2,3 Blindness is costly often trabeculectomy, is indicated, which is associated
to individuals and society.4 Although most people with with potentially serious complications.7
glaucoma do not become blind, many have substantially Surgical lens extraction, as used in managing
impaired quality of life due to restricted peripheral vision age-related cataract, is an alternative approach for the
and the need for long-term treatment.5 Glaucoma has two management of primary angle-closure glaucoma.6,8
subtypes, open angle and angle closure, in which the Age-related growth of the lens plays a major part in the
drainage pathway (trabecular meshwork at the anterior mechanisms leading to primary angle-closure glaucoma,
chamber angle) is blocked or not, respectively.6 Although and lens extraction is used routinely in patients with
primary open-angle glaucoma is more common, primary coexisting cataract. However, the efficacy and safety of
angle-closure glaucoma is more severe and more likely to this treatment in people with primary angle-closure
result in irreversible blindness if not properly treated. glaucoma without cataract has not been fully assessed.9 If
Early and effective interventions are important. lens extraction could control the condition, the need for
Research in context
Evidence before this study after acute attacks of angle closure), lens extraction after other
We identified aCochranesystematic reviewthat had not found any treatments had failed to controlthedisease,or both.
relevant randomised controlledtrials.We also searchedthe
Added value of this study
National ResearchRegister,CurrentControlledTrials,
This large multicentre randomised controlled trial provides
ClinicalTrials.gov, theWHO InternationalClinicalTrial Registry,
evidence supporting the use of initial clear-lens extraction as a
MEDLINE,Embase,the ScienceCitation Index, Biosis, CENTRAL,
first-line intervention for primary angle-closure glaucoma and
and ScopusIn Presswith the following terms: “glaucoma”, “angle-
primary angle closure with high intraocular pressure.
closure”, “PACG”,“phacoemulsification”, “lens extraction”, and
“lens removal” for paperspublished from inceptionuntil June30, Implications of all the available evidence
2007.We updatedour searchesduringthe trial, with the last being Laser peripheral iridotomy as the initial treatment for
donein January, 2016.We found notrials assessing early (clear) angle-closure glaucoma should be reconsidered. This study
lens extraction asthe primarytreatment for chronic primary provides robust evidence that initial clear-lens extraction is
angle-closure glaucoma.We identified clinical trials that had associated with better clinical and patient-reported outcomes,
assessed interventions for primary angle-closureglaucoma,but and that this approach is likely to be cost-effective in a publicly
they studieddifferent populations (eg, patients with cataractor funded health system.
medications and subsequent glaucoma surgery should pressure greater than 21 mm Hg on at least one occasion.
be reduced. Furthermore, clear-lens extraction could help Patients with symptomatic cataract, advanced glaucoma
to maintain good visual acuity and improve quality of life (mean deviation worse than –15 dB or cup-to-disc ratio
by correcting hypermetropic refractive error, which ≥0·9), or previous acute angle-closure attack or who had
frequently affects these patients, and reduce the need for undergone previous laser or ocular surgery were excluded.
wearing spectacles.10 Weighing against initial clear-lens An ophthalmologist identified eligible patients and those
extraction, though, are the potential risks of severe informed about the study were noted in a log book. People
complications associated with intraocular surgery. willing to participate completed clinical measurements
We did a multicentre, randomised, controlled trial to and study questionnaires at baseline.
assess the efficacy, safety, and cost-effectiveness of The study adhered to the tenets of the Declaration of
clear-lens extraction compared with laser peripheral Helsinki and was approved by local institutional review
iridotomy and topical medical treatment as first-line boards. Study participants provided written informed
therapy in people with newly diagnosed primary angle consent. An independent data monitoring committee
closure with raised intraocular pressure or primary and an independent trial steering committee provided
angle-closure glaucoma (ie, those at the highest risk of oversight.
visual loss). We tested the hypothesis that initial clear-
lens extraction would be associated with better quality of Randomisation and masking
life, lower intraocular pressure, and less need for The randomisation schedule was created with a web-based
glaucoma surgery at 36 months than standard care. application at the Centre for Healthcare Randomised Trials,
The protocol of the study has been published.11 University of Aberdeen, Aberdeen, UK. The randomisation
algorithm used sex, centre, ethnic origin (Chinese or
Methods non-Chinese), diagnosis, and one or both eyes suitable
Study design and participants for treatment as minimisation covariates.12 Outcomes
We did this multicentre, comparative effectiveness, assessors were masked when possible, as described later.
randomised, controlled trial in 30 hospital eye services in Patients were assigned in equal proportions to treatment
five countries: Australia (one hospital), mainland China with clear-lens extraction or laser peripheral iridotomy and
(one), Hong Kong (two), Malaysia (two), Singapore (two), topical medical treatment (standard care). Enrolment and
and the UK (22). randomisation was done by the local ophthalmologists.
We recruited patients who were phakic, aged 50 years or
older (to limit the effect of loss of accommodation Procedures
associated with clear-lens extraction), and had newly Topical medications started at the time of diagnosis were
diagnosed primary angle closure with intraocular pressure continued and the interventions were performed within
30 mm Hg or greater or primary angle-closure glaucoma. 60 days of randomisation. If both of a patient’s eyes were
Primary angle closure was defined as iridotrabecular suitable for treatment, they were treated in the same way
contact, either appositional or synechial, of at least 180° but data for the eye with more severe disease was used in
on gonioscopy, and primary angle-closure glaucoma the analysis of eye-level outcomes. If only one eye was
as reproducible glaucomatous visual field defects, suitable for treatment, the other was managed according
glaucomatous optic neuropathy, or both, and intraocular to the clinical judgment of the ophthalmologist.
the decision to submit for publication. Table 2: Patient-reported and clinical primary endpoints
Results
Of 805 patients assessed, 250 did not meet the inclusion the ceiling ratios of £20 000 and £30 000 per QALY
criteria and 136 declined to participate, meaning that gained, the probabilities of cost-effectiveness were 0⋅671
419 individuals were recruited between Jan 8, 2009, and and 0⋅776, respectively. The analysis based on multiple
Dec 28, 2011. 208 were assigned to undergo clear-lens imputation data gave an estimated incremental cost of
extraction and 211 to receive standard care (figure 1). £844 (95% CI 551–1124) for initial clear-lens extraction
155 (37%) participants had primary angle closure and (£2411 vs £1567 for standard care), for an incremental
263 (67%) had primary angle-closure glaucoma. 128 (31%) QALY gain of 0⋅100 (2⋅542 vs 2⋅442, 95% CI 0⋅016–0⋅193).
participants were of Chinese and 291 (69%) were of The estimated ICER after this analysis was £8430, and
non-Chinese ethnicity. In 152 (36%) participants both the probability of clear-lens extraction being cost-effective
eyes were suitable for treatment. One participant was 0⋅885 at a ceiling ratio of £20 000 per QALY and
randomly assigned to clear-lens extraction was classified 0⋅940 at the ceiling ratio of £30 000 per QALY.
as a crossover. 409 (98%) of patients were included in the NEI-VFQ-25 and Glaucoma Utility Index scores at
intention-to-treat analysis (figure 1), with the final visit 36 months were significantly higher in the clear-lens
being held in December, 2014. Baseline characteristics extraction group than in the standard care group (both
were similar in the two treatment groups (table 1). Among p<0·0001, table 3). Significantly fewer participants in the
participants who underwent clear-lens extraction, 18 (9%) clear-lens extraction group needed any treatment to control
had viscosynechiolysis associated with the surgical intraocular pressure (p<0·0001) and fewer needed
procedure. In the standard care group, ten (5%) underwent glaucoma medications than patients who received standard
laser iridoplasty. care (table 3). The most common type of medication used
EQ-5D scores and intraocular pressure at 36 months was prostaglandin analogue (41 [20%] patients in the clear-
significantly favoured the clear-lens extraction group lens extraction group and 105 [50%] in the standard care
(table 2). The economic analysis was based on 179 of group) then β blockers (14 [7%] and 54 [26%]). Degrees of
285 participants recruited in UK centres for whom appositional and synechial angle closure were not reported
complete cost and QALY data were available. Of these, in most participants, but the available data suggest that
93 were assigned to undergo clear-lens extraction and neither differed significantly between groups (odds ratio
86 to receive standard care. Mean adjusted NHS costs 0·78, 95% CI 0·187–3·250 and 0·565, 0·257–1·242,
were higher with initial clear-lens extraction than with respectively). Visual field severity at 36 months was similar
standard care (£2467 vs £1486), resulting in a mean in the two treatment groups (table 3) and had worsened in
incremental cost of £981 (95% CI 612–1317). The 24 participants in the clear-lens extraction group and
corresponding mean QALYs were also higher in the 30 individuals in the laser iridotomy group (odds ratio 0·77,
clear-lens extraction group (2⋅602 vs 2⋅533), resulting in 95% CI 0·392–1·511).
a mean incremental QALY gain of 0⋅069 (–0⋅017 to There were no serious adverse events (table 4). 75 patients
0⋅159). The ICER, therefore, was £14 284 per QALY (25 in the clear-lens extraction group and 50 in the standard
gained for clear-lens extraction versus standard care. At care group) had at least one complication. Posterior
A B
Overall
Ethnicity 0·048 (99% CI –0·055 to 0·151), p=0·23 –1·01 (99% CI –3·32 to 1·31), p=0·26
Non-Chinese 291/419 291/419
Chinese 128/419 128/419
Diagnosis 0·071 (99% CI –0·028 to 0·169), p=0·07 –0·64 (99% CI –2·85 to 1·57), p=0·46
PAC 155/418 155/418
PACG 263/418 263/418
Eye disease 0·077 (99% CI –0·022 to 0·176), p=0·05 –1·00 (99% CI –1·20 to 3·20), p=0·24
Unilateral 266/418 266/418
Bilateral 152/418 152/418
Visual acuity 0·038 (99% CI –0·142 to 0·066), p=0·34 –0·75 (99% CI –1·57 to 3·07), p=0·40
<85 ETDRS letters 290/410 290/410
≥85 ETDRS letters 120/410 120/410
Favours standard care Favours clear-lens extraction Favours clear-lens extraction Favours standard care
Figure2: Mean diff erences in subgroup outcomes between clear-lens extraction and standard care
(A)Quality of life scores onthe EuropeanQuality of Life-5 Dimensions questionnaire. (B) Intraocular pressure. Reddotted vertical line indicates overall difference between
clear-lens extraction and standard care. PAC=primary angle closure. PACG=primary angle-closure glaucoma. ETDRS=EarlyTreatment Diabetic Retinopathy Study chart.
if surgery were performed by less experienced surgeons. masked from participants, nor could the clinical outcome
The number of participants needing further surgery to assessments for gonioscopy or complications. A large
manage complications at any timepoint was similar in proportion of gonioscopic data was not reported.
the two groups. Furthermore, 12 individuals originally Definitions of complications, such as inflammation and
assigned to laser peripheral iridotomy needed cataract spikes in intraocular pressure, were not standardised.
surgery for visual reasons, which suggests that many There might have been a difference between Chinese and
people treated with this approach will be at risk of future non-Chinese populations that we did not detect and,
cataract extraction. The lack of difference between groups therefore, the generalisability of our results to Asian
for central corneal thickness also supported the relative populations should be explored further.
safety of clear-lens extraction as first-line treatment. Although one good-quality trial might not be enough to
Within-trial cost-effectiveness data showed that change policy, the consistent superiority of clear-lens
clear-lens extraction was associated with increased mean extraction in terms of patient-reported and clinical benefits
cost to the NHS and increased mean QALYs at 3 years. At and the absence of serious safety issues provide strong
a ceiling willingness-to-pay ratio of £20 000 per QALY support for considering this approach as the first-line
gained, the probability of early clear-lens extraction being treatment for individuals with primary angle-closure
cost-effective is 67% based on complete case data, or disease. The results are consistent with those in previously
89% based on the multiple imputation analysis. The latter published case series. Longer-term follow-up of visual
analysis suggested that patients with missing economic function and visual field progression could be useful,
data were predicted to have poorer outcomes for quality owing to the slow progression of this disease when treated.
of life than those with complete data, particularly in the Overall, our results indicate that clear-lens extraction
standard care group (as well as higher costs in this group), should be considered the initial treatment for primary
leading to a slightly more favourable ICER for early clear- angle-closure glaucoma and primary angle closure with
lens extraction. The incremental cost associated with increased intraocular pressure.
early clear-lens extraction is driven by increased initial Contributors
procedure costs (£1229 vs £181), but these are partly offset AA-B, JB, CR, and JN conceived and designed the trial. AA-B and JB
over follow-up by cost savings associated with reduced were the chief investigators and oversaw the trial throughout. CC was
the trial coordinator. CR and DC planned and did the statistical analysis.
need for subsequent procedures and medications. The GS and MJ planned and did the health economic analysis. All authors
cost-effectiveness of clear-lens extraction might, therefore, contributed to the interpretation of data, drafting of the report, and
improve further in the long term if this trend continues. decided on its content. All authors approved the final version.
This trial has several strengths, including its pragmatic Trial steering committee
design, the large sample with low attrition, the involvement Independent members: Richard Wormald (chair), Roger Hitchings,
of centres in the UK and Asia, the randomisation process, Colm O’Brien, Russell Young. Non-independent members: Augusto
Azuara-Blanco, Jennifer Burr, John Norrie, Craig Ramsay, Luke Vale,
and masking of the clinical assessments of intraocular Paul Foster, Winifred Nolan, David S Friedman, David Wong, Tin Aung,
pressure, visual acuity, and visual fields, which kept the Paul Chew, Jovina See, Jimmy Lai, Gus Gazzard, Jemaima Che-Hamzah.
potential risk of bias to a minimum. This study also had Data monitoring committee
some limitations. The surgical treatments could not be David Garway-Heath (chair), David Crabb, Baljean Dhillon.