Clinical and technical evidence

A literature search was carried out for this briefing in line with the interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

Five observational studies are summarised in this briefing, including 166 patients. One study included patients with DMO and a cohort of healthy volunteers (Sahni 2017). These participants used Noctura 400 Sleep Mask, which exposes both eyes to light. In a second study, patients with DMO with either advanced non-proliferative or early proliferative DR also used Noctura 400 Sleep Mask (Kuchynka 2017). A third study, for which only an abstract was available, was a retrospective review of patients with diabetes (some of whom had no retinopathy) who used the mask over a period of 16 months (Ng 2017).

An earlier version of the mask, known as the Arden mask, was used in 2 further studies with a total of 46 patients (Arden 2010; Arden 2011). Unlike Noctura 400 Sleep Mask, the Arden mask only exposed 1 eye to light, so patients in these studies acted as their own controls (that is, the eye exposed to the light through the mask was assessed against the other eye that was not exposed). Aside from this difference, the Arden mask appears to be functionally very similar to the Noctura 400 and so this evidence was judged to be relevant.

Table 2 summarises the clinical evidence, as well as its strengths and limitations. Two of the studies using Noctura 400 Sleep Mask reported some small improvements in retinal thickness (Sahni 2017) and best corrected visual acuity (Kuchynka 2017). These studies focused largely on safety and usability of the mask and no other changes in vision were observed. Some adverse events associated with the mask were reported, including eye soreness and eyelid swelling, itchy and dry eyes and general discomfort when wearing the mask. The third study reported limited data as it was an abstract only, but showed no progression of retinopathy (measured through retinopathy grading and best corrected visual acuity) over the course of follow-up.

The 2 studies using the Arden mask also reported small improvements in retinal function following mask use. One study showed a reduction in thickness of the zone of maximal oedema, a reduction in the number of intraretinal cysts and improvements in visual acuity in treated versus untreated eyes in patients with mild DR (Arden 2011). The other study showed an improvement in colour vision (tritan thresholds) associated with the Arden mask (Arden 2010). Neither study reported any adverse events or complications.

Overall assessment of the evidence

The evidence for Noctura 400 Sleep Mask comes from 5 studies, 2 of which used a predecessor of the mask. The evidence gives a limited indication that Noctura 400 Sleep Mask may provide beneficial effects to patients with mild, non-proliferative DR and DMO, or prevent progression of these conditions in patients with diabetes who do not currently have DR or DMO. No evidence has been presented for treating patients with diabetes with advanced (proliferative) DR or DMO.

However, the study populations were limited, both in the number of patients included and the severity of disease. Only 1 study compared outcomes in patients using the mask against a group of healthy controls, and none compared treatment with the mask against other available therapies. The outcomes in the study were mainly measures of safety and usability and intermediate or proxy outcomes for retinal function, such as retinal thickness. This meant that the reported results were largely limited to the technical success of the mask, rather than any more patient-centred outcomes.

A high-quality randomised control trial is needed to support the use of this technology. Studies are also needed on the cost effectiveness of the technology compared with standard care, particularly for laser photocoagulation.

Table 2 Summary of included studies

Arden 2010

Study design and population

A 2-centre observational study in Germany and Turkey, including 12 patients with mild to moderate non-proliferative DR.

Intervention(s) and follow-up

Arden mask (1 eye treated, 1 eye used as control), worn every night for 1 year in the German patients and 3 months in the Turkish patients (patient numbers in each country not reported).

Results

Retinal function

Colour vision: In 9/10 patients there was improvement in tritan thresholds in treated versus untreated eyes: Mean (SD) in micrometres at baseline, 6.7 (1.9) versus 8.8 (6.0); mean (SD) at study end, 6.5 (1.4) versus 10.0 (7.0); p=0.03.

Protan thresholds remained unchanged throughout the study in treated versus untreated eyes: Mean (SD) at baseline, 2.9 (0.67) versus 3.19 (1.05); mean (SD) at study end, 2.95 (0.56) versus 3.17 (1.04); p=0.359.

Dark retinal anomalies (microaneurysms, dot haemorrhages): Fundus photography showed that areas of dark anomalies (as a proportion of total retinal area), indicating microaneurysms and/or dot haemorrhages, decreased in treated eyes over the course of the trial: Mean (SE) at baseline, 1.31% (0.73%); at study end, 0.73% (0.45%); p=0.07 (n=11). Areas of dark anomalies increased in untreated eyes: Mean (SE) at baseline, 0.69% (0.41%); at study end, 0.92% (0.37%). The binomial probability that treatment was ineffectual and these differences were because of chance was calculated as 0.01.

Safety/patient experience

No patients reported any ocular or systemic complications during the study.

Strengths and limitations

Strengths: Use of statistical analysis.

Limitations: Small number of patients; inconsistent follow-up, no true control group (non-treated eye used).

Arden 2011

Study design and population

A single-centre observational study in the UK, including 34 patients with mild non-proliferative DR and early untreated DMO.

Intervention(s) and follow-up

Arden mask (1 eye treated, 1 eye used as control), worn every night for 6 months.

Follow-up after 3 and 6 months of wear.

Results

Retinal function

Thickness: Swelling increased in 8 and 16 treated and untreated eyes respectively, and decreased in 26 and 18 eyes respectively.

Mean (SD) macular cube thickness for treated eyes and untreated eyes at baseline was 282 (51) micrometres and 278 (25) micrometres respectively, and changed insignificantly after 6 months of wear to 279±16 micrometres and 279±22 micrometres respectively. The authors attribute this to the fact that DMO was localised.

There were significant changes in the thickness of the zone of maximal oedema over 6 months. Months 1 to 6: mean (SD), 12.0 (23.8) micrometres; p=0.01. Months 1 to 3: mean (SD), 9.1 (21.9) micrometres; p=0.01. Months 3 to 6: mean (SD), 7.6 (22.7) micrometres; p=0.04. The difference between mean change in the treated eye and mean change in the untreated eye was −13 (95% CI, −17.5 to −7.2) micrometres p=0.052, suggesting that thickness in the treated eyes reduced to a greater extent than in the untreated eyes.

There were no significant changes in the thickness of the central subfield zone over 6 months. Months 1 to 6: mean (SD), 5.9 (34.2) micrometres; p=0.18. Months 1 to 3: mean (SD), 6.3 (28.9) micrometres; p=0.13. Months 3 to 6: mean (SD), 7.7 (47.8) micrometres; p=0.20. However, the difference between mean change in the treated eye and mean change in the untreated eye was significant: −11 (95% CI, −14.7 to −6.9) micrometres ; p=0.026). Cysts: 28 treated eyes and 9 untreated eyes had intraretinal cysts at baseline. After 6 months of wear, 9 of the cysts in treated eyes had disappeared or reduced considerably, while the number of untreated eyes with cysts increased to 20.

Visual acuity: At baseline, there was no statistically significant difference in acuity (ETDRS letters) between treated eyes and untreated eyes: Mean (SD), 78.04 (7.80) and 76.52 (8.10) respectively; p=0.39. However, after 6 months there was a statistically significant difference in acuity, indicating improvements in the treated eyes: mean (SD), 80.18 (5.32) and 74.59 (8.52) respectively; p=0.001. This is just below the threshold for clinical significance (5 letters).

Contrast sensitivity: At baseline, there was no statistically significant difference for CS (Pelli-Robson) between treated eyes and untreated eyes: mean (SD), 31.71 (5.38) and 30.16 (4.71) respectively; p=0.214. However, after 6 months there was a non-statistically significant difference in CS, indicating improvements in the treated eyes: Mean (SD), 33.14 (4.57) and 30.50 (4.95) respectively; p=0.054.

BCVA: 10/13 patients with BCVA <80 in the treated eyes at baseline improved by 5 letters or more. For untreated eyes, 3/9 patients with BCVA <80 at baseline improved, while 6 remained unchanged and 3 deteriorated by >3 letters.

Safety/patient experience

No patients reported any adverse events, mood alteration or difficulties with wearing the mask or sleeping.

Strengths and limitations

Strengths: Use of statistical analysis.

Limitations: Low patient numbers; masks had a high failure rate (light not administered continuously); untreated eyes used as the control and were not equal in all respects at baseline to treated eyes (number of cysts and thickness of ETDRS zone 1).

Kuchynka 2017

Study design and population

A single-centre, prospective, open-label clinical trial conducted in the Czech Republic, including 45 patients with DMO.

Intervention(s) and follow-up

Noctura 400 Sleep Mask, worn every night for 6 months.

Patients received new masks monthly. Patients had either advanced non-proliferative or early proliferative DR.

Results

Safety/patient experience

Adverse events: No major safety issues relating to using the mask were identified. However, a minority of patients reported mask discomfort, rubbing of straps and other design-related issues. Results from a patient questionnaire showed that 23% found the mask uncomfortable, 14% reported sleeping difficulties that were beyond what they had experienced before mask wear, 14% found the mask light intrusive, and 9% experienced new trouble falling asleep.

Three (7%) patients left the trial early because of light intolerance or mask wear intolerance. An additional 3 patients left the trial because of severe adverse events that were judged not to be related to the mask (4 further patients left the trial for unknown reasons).

Treatment compliance: Most patients wore the mask nightly, but mask wear generally fell short of the 8 hours available. Mean nightly mask use was 4.96 (SD=2.97) hours. Two patients wore the mask for mean periods of less than 2 hours per night.There was a downward trend in mask wear after 3 months of use, although this was not statistically significant.

Patient experience: After 6 months of mask wear, 26% patients felt that they were not willing to continue wearing the mask and 23% were unsure whether they were willing to continue wearing the mask.

Retinal function

BCVA improved following 13% to 14% over 6 months of mask wear. The results of the OCT of CST over 6 months were mixed with 38.8% of eyes examined showing improvement, 29.9% remaining the same and 31.1% deteriorating. Overall there was thinning of the CST from before the trial by 12 micrometres and 27 micrometres for the left and right eye respectively.

Strengths and limitations

Strengths: -

Limitations: One of the authors acts as a medical and scientific adviser to the company; patients received new masks monthly, but it is generally assumed that masks are changed every 12 weeks in clinical practice (in line with the company's instructions); no control group.

Ng 2017

Study design and population

A retrospective review of clinical data conducted in the UK, including 15 patients with diabetes with at least 1 year of follow-up.

Intervention(s) and follow-up

Noctura 400 Sleep Mask, worn for an average of around 18 months. Patients self-funded use of the mask.

Results

Retinal function

BCVA: 74±18 and 67±28 in the right and left eyes respectively, at baseline. After 8 months of using the mask, BCVA was 77±15 and at month 16 it was 72±19.

Retinopathy grading: ranged from R0M0 (no retinopathy) to R3sM1P1 (stable treated retinopathy) at baseline, and remained stable throughout the study.

Safety/patient experience

No significant safety signal was detected. Mean nightly duration of mask use was 5 hours 11 minutes per night (65% of the 8-hour illumination provided by the mask).

Strengths and limitations

Strengths: Use of real-world data.

Limitations: Low patient numbers; abstract only presented so more detailed results (for example, statistical analysis) are not available; possibility of response bias as patients in a self-funded pilot of the mask may be more motivated and, therefore, more likely to show compliance.

Sahni 2017

Study size, design and location

A single-centre, prospective, longitudinal safety study conducted in the UK, including 60 patients.

15 patients with DMO (group C) and 45 healthy controls (group A, 21 patients aged 18–30 years; group B, 24 patients aged 50–70 years).

Intervention and comparator(s)

OLED sleep mask (Noctura 400 Sleep Mask, Polyphotonix Medical Ltd), worn every night for 3 months.

Key outcomes

Retinal function

Thickness: In the DMO group, there were significant changes in mean thicknesses per 1 micrometre change in baseline thickness for CST and maxST at 3 and 4 months. CST: month 3, −0.77 micrometres, p<0.001; month 4, −0.79 micrometres p<0.001. MaxST: month 3, −0.44 micrometres ,p=0.05; month 4, −0.62 micrometres, p<0.01.Cysts: 10/15 (67%) patients in the DMO group showed a reduction/clearance of cysts in the maximum pathology subfield.

Other measures of vision: No significant change was detected in DMO patients after 3 months of wear in CCT, BCVA, mfERG or EOG (change measured at 4 months). However, a small, but statistically significant change was observed for CS (letters), p=0.01.

Safety/patient experience

Loss of concentration/alertness caused by sleep deprivation: Performance in a psychomotor vigilance test (reaction time and number of lapses) deteriorated in the healthy control groups after 3 months. Patients in the DMO group had poorer performance at baseline, but showed no statistically significant change during the study.

Psychological wellbeing: GHQ12 scores in all groups worsened in after 3 months (Group A −28.0%,p=0.01; group B −21.2%, p=0.03; group C −12.7%, p<0.05).

Adverse events: 75% participants in the healthy control groups and 40% in the DMO group who wore the sleep mask for ≥1 month reported ≥1 adverse event. The most common were: eye soreness (n=5; 5 healthy controls, 0 DMO patients); eyelid swelling (n=4; 4 healthy controls, 0 DMO patients); itchy eyes (n=4; 3 healthy controls, 1 DMO patient); and dry eyes (n=3; 2 healthy controls, 1 DMO patient). Events were mostly attributed to the fabric mask. One serious adverse event occurred that was deemed unlikely to be related to the mask.

Sleep disturbance: 6 patients in the healthy control groups withdrew before study completion because of light intolerance and sleep disturbance caused by the mask. No patients in the DMO group withdrew for this reason (although 2 did withdraw: 1 following an unrelated medical event, and 1 for no reason provided).

Strengths and limitations

Strengths: Prospective study design; use of statistical analysis.

Limitations: Relatively low patient numbers; no comparison with standard care for DMO, or with no mask wear; high number of study withdrawals (16/60).

Abbreviations: BCVA, best corrected visual acuity; CCT, Cambridge colour vision test; CS, contrast sensitivity; CST, central subfield thickness; DMO, diabetic macular oedema; DR, diabetic retinopathy; EDTRS, Early Treatment Diabetic Retinopathy Study; EOG, electrooculogram; GHQ12, General Health Questionnaire of psychological wellbeing; maxST, mean thickness of subfield with maximal pathology; mfERG, multifocal electroretinogram; OLED, organic light-emitting diode; SD, standard deviation; SE, standard error.

Recent and ongoing studies

A search on clinicaltrials.gov identified the following ongoing trial:

  • Noctura400 treatment for diabetic retinopathy (CANDLE). Multicentre, interventional, randomised, parallel assignment, open-label trial (patients receiving ranibizumab only versus experimental: Noctura 400 with ranibizumab) ClinicalTrials.gov identifier: NCT02207712. Status: Ongoing. Device: Noctura 400.

One ongoing study was identified through citation searching during study selection: