Diabetic Macular Oedema and Diode Subthreshold Micropulse Laser (DIAMONDS): A Pragmatic, Multicentre, Allocation Concealed, Prospective, Randomised, Non-inferiority Double-masked Trial
Diabetic Macular Oedema and Diode Subthreshold Micropulse Laser (DIAMONDS): A Pragmatic, Multicentre, Allocation Concealed, Prospective, Randomised, Non-inferiority Double-masked Trial
DIAMONDS is a Randomised Controlled Clinical Trial that is being carried out in the UK to determine the clinical effectiveness and cost-effectiveness of micropulse laser, compared with standard laser, for the treatment of diabetic macular oedema.
Background
Laser has been the treatment of choice for people with Diabetic Macular Oedema (DMO) since its beneficial effects were demonstrated by the ETDRS in 1985 [1]. In a recent appraisal [2], The National Institute of Health and Care Excellence (NICE) recommended the continued use of standard threshold laser to treat people with DMO and central retinal thickness of < 400 microns because laser treatment is clinically effective and cost-effective in this patient group. However, conventional laser treatment may have side effects including paracentral scotomas (areas around the central vision in which patients do not see; these may affect the ability to read and drive), enlargement of the laser scar over time with potential visual loss, reduced colour vision and epiretinal membrane/subretinal fibrosis. If conventional laser is accidentally applied to the centre of the retina, profound visual loss would ensue.
Small case series and small randomised trials suggested DSML has comparable or superior efficacy but may have less side effects than conventional laser. Thus, Lavinsky and collaborators [3], in a trial where participants were randomised to receive standard threshold laser (n=42), normal density DSML (n=39) or high density DSML (n=42) showed superiority of the latter with regards to visual acuity improvement and reduced retinal thickness at 12 months follow-up [3]. In a randomised trial with 50 patients with DMO, Vujosevic and colleagues [4] found no differences in terms of visual acuity or central retinal thickness between standard threshold laser and DSML, but a statistically significant increased retinal sensitivity, as measured with microperimetry was achieved following DSML with no laser scars present in the retina, as disclosed by fundus examination and autofluorescence images at 12 months follow-up [4]. Similarly, randomised trials by Kumar and associates [5], Figueira and coworkers [6] and Laursen and associates [7] included 20, 53 and 16 patients respectively, and a follow-up duration of 18 weeks, 12 months and 5 months respectively, found no differences in visual acuity and central retinal thickness between conventional threshold laser and diode subthreshold laser. A recently published review [8] concluded that available data suggested similar or superior efficacy of DSML (when compared with standard threshold laser) with less or no retinal damage. It is possible that DSML may allow for a more standard delivery of treatment to all patients given that it is applied to the entire macular area in a confluent manner, rather than to zones that the treating ophthalmologist may perceive as thickened or to areas of suspected leakage or extrafoveal ischaemia based on fluorescein angiography, reducing/minimising possible variability in the results obtained.
Although published data suggest that DSML may be superior to conventional threshold laser, current knowledge is based on results of randomised trials with small numbers of patients followed for no more than 12 months. Furthermore, important outcomes such as cost-effectiveness of the treatment, effects of the laser on the visual field and patient reported outcomes were not evaluated in these previous studies. Stronger evidence is required to support policy-making and investment decisions.
Rationale for the Study
Diabetic Macular Oedema (DMO) is the most common cause of irreversible blindness among people with diabetes mellitus and Diabetic Retinopathy (DR). It was estimated that in England alone, the prevalence of DMO in one or both eyes was 166,325 or 7% of all people with diabetes, of whom 40% had some reduction in visual acuity [9]. DMO represents accumulation of fluid in the macula, the area responsible for central detailed vision. As fluid accumulates, visual loss ensues. NICE [2] has recommended the use of standard threshold laser to treat people with DMO and retinal thickness of < 400 microns, as measured using an imaging technique called optical coherence tomography (OCT), because it dominated when compared with anti-VEGF (ranibizumab or aflibercept) treatment. Randomised trials demonstrated the efficacy of laser to prevent visual loss in people with DMO. The ETDRS found a 50% reduction in visual loss 3 years following laser and recent randomised trials have shown that as much as 35% of patients experience an improvement in vision of at least 10 ETDRS letters following laser. Furthermore, approximately 50% of people undergoing anti-VEGF therapy still require macular laser within the first two years of treatment.
Laser thus remains an effective option for people with DR and DMO. However, side effects of laser include paracentral scotomas, which may affect the ability to undertake tasks that require precise near vision, such as reading, and the ability to drive; enlargement of the laser scar over time with a potentially associated visual loss; reduced colour vision; epiretinal membrane and subretinal fibrosis formation, among others. If the laser is accidentally applied to the centre of the macula, marked central visual loss would occur.
Standard threshold laser is performed using a continuous wave laser that produces a visible burn (a "grey" mark) in the retina. Retinal cells at the site of the burn are killed. The laser energy is absorbed by one of the layers of the retina, the retinal pigment epithelium (RPE), and converted into heat. Although the mechanisms of action of conventional threshold laser are not completely understood, it is believed that it has its effect by acting upon still viable RPE cells around the site of the laser spot. As heat spreads by conduction, a damaging effect in retinal layers overlying the RPE, including the photoreceptors (visual cells) may also occur. Standard laser requires considerable expertise by the clinician in order to identify areas involved to which the laser should be aimed. Affected areas requiring laser may be determined by slit-lamp biomicroscopy using a contact lens, with the help of optical coherence tomography and/or fluorescein angiography.
Recently, DSML was introduced. In DSML, a series of repetitive short laser pulses are applied ("micropulse"), instead of a continuous wave emission, each separated by a long off-time which reduces the increased temperature in the tissue that follows conventional laser. In this manner, a sublethal effect on the RPE is achieved with preservation of the overlying neurosensory retina and visual cells. Small case series and randomised trials including relatively small number of patients have shown that subthreshold tissue-sparing micropulse laser may have comparable or higher efficacy than standard laser, even in the absence of a visible burn, with reduced side effects. This laser may be easier to deliver because it is applied to the entire macular area and, thus, may be less dependent on surgeon"s skills.
An adequately powered randomised trial comparing standard laser versus DSML, such as DIAMONDS, would benefit patients by providing strong evidence to advise whether one or the other should be preferred. This is important because it is likely that laser will remain an option to patients with DMO for many years to come. Furthermore, because DIAMONDS is a pragmatic randomised trial conducted in a National Health Service (NHS) setting, its findings are likely to be of relevance and practical use to the NHS for some time. Several measures of visual function will be obtained from the DIAMONDS trial to gain further knowledge on the effects of conventional and DSML in the retina. The knowledge gained might also be applicable to other retinal disorders.
Research Hypothesis
The hypothesis is that DSML is non-inferior to standard laser for the treatment of patients with DMO and a central subfield thickness of < 400 microns.
Study Aim
DIAMONDS aims to determine the clinical effectiveness and cost-effectiveness of DSML compared with standard laser for the treatment of people with DMO and central retinal subfield thickness of < 400 microns, for which laser treatment is currently recommended by NICE [2].
Primary Objective:
The primary objective of DIAMONDS is to determine whether DSML is as good or superior to standard laser at improving or preserving vision at 24 months following treatment in patients with DMO.
Secondary Objectives:
The secondary objectives of DIAMONDS are to determine whether DSML is as good or superior to standard laser at improving or preserving binocular vision and visual field, reducing / clearing DMO, allowing treated patients to achieve driving standards and improving their health and visual related quality of life at 24 months following treatment. The relative cost-effectiveness of DSML when compared with standard laser will also be evaluated, as well as side effects of these treatments, number of laser treatments required and use of additional treatments (other than laser) for both, DSML and standard laser.
Patient Recruitment
The DIAMONDS trial requires 266 participants to be recruited. Using an electronic database from a typical UK centre serving a population of 600,000 we confirmed that 75 patients with DMO of < 400 microns were evaluated at this centre in one year. Based on this, we estimate that the centres participating in DIAMONDS, which serve a total population of more than 6,000,000 people, will assess at least 900 people with DMO < 400 microns during the recruitment period. Based on our clinical experience, we anticipate that most of these patients will be eligible to participate in DIAMONDS. Estimating, conservatively, that 50% of eligible patients will be identified in time for treatment and that 70% of those asked if they wish to participate will agree, the research group should be able to recruit the targeted 266 individuals in approximately 15 months, with each centre recruiting an average of 2 participants per month.
Potential participants will be identified through patient electronic databases at each of the participating centres, through referrals to Hospital Eye Services or while in the clinic. If identified through search of electronic databases (e.g. Medisoft, an electronic database in routine use in Ophthalmic Clinics) or through letters of referral to the Hospital Eye Services the potential participants may be approached to participate in the study by phone or via an invitation letter. When approached by phone, the potential participants will be informed about the study before they come to their Hospital appointment; if willingness to participate is demonstrated, a patient information leaflet will be sent to them prior to the clinical appointment. Then, at their clinical appointment and if agreeable to participate, informed consent will be obtained and the patient will be recruited in the study while in clinic. If they are approached by letter, a letter of invitation to participate in the study and a patient information leaflet will be provided to the potential participant prior to their clinical appointment. Then, as above, when the patient comes to their clinical appointment, if willing to participate, they will be consented and enrolled in the study. Under the above circumstances, potential participants will have a minimum of 24 hours to decide whether or not they wish to participate in the study.
Potential participants may be also identified while in the clinic. In this case, information about the study will be given there and then, including a patient information leaflet. Under these circumstances, patients will be asked whether they wish to have time to think about their participation in the study once information has been provided to them and once they have had time to ask questions about it. It is envisaged that some patients would like to have time to think about their potential participation in the study, in which case a further visit will be organise for them; if at this further visit they are willing to participate, then they will be recruited. It is envisaged also that some patients may not wish to delay their laser treatment (i.e. having 24 hours to think about it and then having to arrange an additional visit) but may wish to participate in the study. Under these circumstances, patients will be given the time they wish to think about the study and, if willing to be recruited on the same day, following informed consent, they will be recruited into the trial.
Pilot Study
An internal pilot study to assess feasibility will be undertaken. This will run until month 4 of the recruitment period, by which time it is expected to have recruited 50 patients. Recruitment feasibility milestones will be as follows. If recruitment rates achieve 75-100% recruitment, we will progress with the trial; if we achieve 50-75% recruitment, we will progress with the trial following review of screening logs and the protocol and after barriers to achieving adequate recruitment are addressed; if we recruit 25-50% of the required number, the trial will be progressed only after screening logs and the protocol are reviewed and following approval by NIHR HTA, additional sites are opened. Should recruitment be <25%, it is not expected the trial will progress. The decision to stop the trial will be made by the TSC and the NIHR HTA.
Randomisation - Treatment Allocation
Once consent has been obtained from patients meeting the eligibility criteria they will be recruited to the study.
On the day the laser procedure is going to be performed, participants will be randomised 1:1 to receive DSML or standard laser using an automated randomisation system. The local ophthalmologist will be the person interacting with the automated randomisation system to generate the random allocation sequence. The laser procedure should be performed within 2 weeks of the baseline visit. If the laser treatment is not being performed on the same day as the baseline visit, eligibility should be confirmed again prior to undertaking the laser treatment.
A minimisation algorithm will be used to ensure balanced allocation of patients across the two treatment groups for the following important prognostic factors: centre, distance BCdVA at presentation [≥ 69 ETDRS letters (Snellen equivalent of ≥ 20/40; logMAR ≥ 0.3); 24-68 ETDRS letters (Snellen equivalent ≤20/50; logMAR 0.4-1.2) previous use of anti-VEGF therapies in the study eye, previous use of macular laser treatment in the study eye.
When a patient is ready to be randomised, the site should access the automated randomisation system and complete all requested information. The randomisation service will assign a unique trial identifier to each patient and issue the treatment allocation, ensuring that each patient"s allocation remains concealed up to the time that it is issued. The randomisation service will confirm randomisation details by email to the site and the NICTU.
The unique trial identifier assigned at the time of randomisation will be used throughout the trial for the purposes of patient identification.
Masking of Treatment Allocation
This randomised trial is designed to be a pragmatic trial so that it"s results would be applicable immediately in a NHS setting. For this reason, ophthalmologists undertaking laser treatments for DMO at each of the participating centres will also deliver the treatment for the trial. Although, for obvious reasons, ophthalmologists delivering the treatment will not be masked with regards to the laser used, every effort will be made to ensure that participants and outcome assessors (e.g. optometrists measuring visual function, photographers/technicians/nurses obtaining OCT images and ophthalmic technicians obtaining visual fields) will be masked to the allocated treatment. Patients will not be informed before, during, and after the laser treatment about which technology of laser was used for their treatment.
Similarly, the investigators obtaining outcome measures will only have access to the CRF booklet (but not to the notes of the patients) which will contain no information with regards to the type of laser the patient had been allocated or received.
Participant Retention and Follow-up
Local study groups, supported by each local PI, will assure that participation in this pragmatic trial will not represent a burden to participants and assure that retention during the 2 year period of the study will be achieved.
Patients will be followed, as per routine standard clinical care, at 4 month intervals (baseline, and at months 4, 8, 12, 16, 20 and 24) for a total of 7 visits.
In order to ensure adequate follow-up of participants, participants will be reminded by telephone, text or call the week prior to the study visit. This will be carried out by either research nurses or by administrative staff at each of the participating centres.
Assessment of safety
The safety of the treatment will be assessed at each visit by noting any complications during or after laser treatment, including self-reported visual disturbances, > 10 letter score ETDRS visual acuity loss, and > 15 letter score ETDRS visual acuity loss. Patients will be asked about reduced colour vision, presence of paracentral scotomas and/or distorsion ("waviness" of straight lines). Although Serious Adverse Events (SAE) related to the study procedures are unlikely to occur, a record will be kept of all SAEs. The NICTU will be responsible for informing the Sponsor and the Research Ethics Committee about any SAEs all study sites about any SAEs. The DMEC will provide information on all SAEs on a routine basis.
Sample Size
The study is powered to demonstrate non-inferiority of DSML with respect to the primary outcome (BCdVA in the study eye at 24 months). The study will have sufficient statistical power to determine superiority of one laser over the other. Furthermore, the study will also have sufficient statistical power to determine equivalence of the lasers.
Based on a mean (standard deviation(SD)) of 0.08 (0.23) log MAR for BCdVA change from baseline for the standard care laser [3] and a permitted maximum difference of 0.1 logMAR (5 ETDRS letters) between groups it is estimated that the trial will require 113 patients per arm at 90% power and 0.05 level of significance. Allowing for a possible 15% dropout rate, which is similar to that observed in other randomised trials on diabetic macular oedema with outcomes determined at 24 months [15, 16], a total of 266 patients will be required for the study.
A permitted maximal difference of 5 ETDRS letters between groups was chosen as the non-inferiority margin because a 5 ETDRS letter difference would not be considered clinically relevant or meaningful to patients (2).
The proposed sample size of 113 per group (which allows for 15% drop out) will be sufficient also to detect a mean difference between lasers of 37.7 microns in central retinal thickness (based on a SD of 86.8 as per [4] and 6.55 in NEI-VFQ based on a SD of 15.1 as per [17]
which are important secondary outcomes on this study. These differences in central retinal thickness and NEI-VFQ scores are both clinically relevant differences [18, 19].
Inclusion Criteria:
Patients with diabetic retinopathy and centre involving DMO, as determined by using spectral domain optical coherence tomography (SD-OCT), in one or both eyes with:
Exclusion Criteria:
Eyes of patients will not be included in the study if: