The purpose of this evidence-based review was to examine the effectiveness and cost-effectiveness of spectral-domain (SD) optical coherence tomography (OCT) in the diagnosis and monitoring of patients with retinal disease, specifically age-related macular degeneration (AMD) and diabetic macular edema (DME). Specifically, the research question addressed was:
What is the sensitivity and specificity of spectral domain OCT relative to the gold standard?
Clinical Need: Target Population and Condition
The incidence of blindness has been increasing worldwide. In Canada, vision loss in those 65 years of age and older is primarily due to AMD, while loss of vision in those 18 years of age and older is mainly due to DME. Both of these conditions are diseases of the retina, which is located at the back of the eye. At the center of the retina is the macula, a 5 mm region that is responsible for what we see in front of us, our ability to detect colour, and fine detail. Damage to the macula gives rise to vision loss, but early detection of asymptomatic disease may lead to the prevention or slowing of the vision loss process.
There are two main types of AMD, ‘dry’ and ‘wet’. Dry AMD is the more prevalent of the two, accounting for approximately 85% of cases and characterized by small deposits of extracellular material called “drusen” that build up in Bruch’s membrane of the eye. Central vision loss is gradual with blurring and eventual colour fading. Wet AMD is a less prevalent condition (15% of all AMD cases) but it accounts for 90% of severe cases. It’s characterized by the appearance of retinal fluid with vision loss due to abnormal blood vessels/leakage within weeks to months of diagnosis. In 2003, the Canadian National Institute for the Blind (CNIB) prevalence estimate for AMD was 1 million Canadians, including approximately 400,000 affected Ontarians. The incidence in 2003 was estimated to be 78,000 new cases in Canada, with approximately one-third of these cases arising in Ontario (n=26,000). Over the next 25 years, the number of new cases is expected to triple.
DME is caused by complications of diabetes mellitus, both Type 1 and Type 2. It is estimated that 1-in-4 persons with diabetes has this condition, though it occurs more frequently among those with type 2 diabetes. The condition is characterized by a swelling of the retina caused by leakage of blood vessels at the back of the eye. In early stages of the disease, vision may still be normal but it can degrade rapidly in later stages. In 2003, the CNIB prevalence estimate for DME was 0.5 million Canadians, with approximately 200,000 Ontarians affected. The incidence of DME is more difficult to ascertain; however, based on an annual incidence rate of 0.8% (for those 20 years of age or older) and the assumption that 1-in-4 persons with diabetes is affected, the incidence of DME in Ontario is estimated to be 21,000 new cases per year.
Optical Coherence Tomography
Prior to the availability of OCT, the standard of care in the diagnosis and/or monitoring of retinal disease was serial testing with fluorescein angiography (FA), biomicroscopy (BM), and stereo-fundus photography (SFP). Each of these is a qualitative measure of disease based on subjective evaluations that are largely dependent on physician expertise. OCT is the first quantitative visual test available for the diagnosis of eye disease. As such, it is allows for a more objective evaluation of the presence/absence of retinal disease and it is the only test that provides a measure of retinal thickness. The technology was developed at the Michigan Institute of Technology (MIT) in 1991 as a real-time imaging modality and is considered comparable to histology. It’s a light-wave based technology producing cross-sectional images with scan rates and resolution parameters that have greatly improved over the last 10 years. It’s also a non-invasive, non-contact visual test that requires just 3 to 5 minutes to assess both eyes.
There are two main types of OCT system, both licensed by Health Canada as class II devices. The original patent was based on a time domain (TD) system (available from 1995) that had an image rate of 100 to 400 scans per second and provided information for a limited view of the retina with a resolution in the range of 10 to 20 μm. The newer system, spectral domain (SD) OCT, has been available since 2006. Improvements with this system include (i) a faster scan speed of approximately 27,000 scans per second; (ii) the ability to scan larger areas of the retina by taking six scans radially-oriented 30 degrees from each other; (iii) increased resolution at 5μm; and (iv) ‘real-time registration,’ which was not previously available with TD.
The increased scan speed of SD systems enables the collection of additional real-time information on larger regions of the retina, thus, reducing the reliance on assumptions required for retinal thickness and volume estimates based on software algorithms. The faster scan speed also eliminates image distortion arising from patient movement (not previously possible with TD), while the improvement in resolution allows for clearer and more distinguishable retinal layers with the possibility of detecting earlier signs of disease. Real-time registration is a new feature of SD that enables the identification of specific anatomical locations on the retina, against which subsequent tests can be evaluated. This is of particular importance in the monitoring of patients. In the evaluation of treatment effects, for example, this enables the same anatomic retinal location to be identified at each visit.
A literature search was performed on February 13, 2009 using Ovid MEDLINE, MEDLINE In-Process & Other Non-Indexed Citations, EMBASE, the Cumulative Index to Nursing & Allied Health Literature (CINAHL), the Cochrane Library, and the International Agency for Health Technology Assessment (INAHTA) for studies published from January 2003 to February 2009. The subject headings and keywords searched included AMD, DME, and OCT (the detailed search strategy can be viewed in Appendix 1). Excluded were case reports, comments, editorials, non-systematic reviews, and letters. Abstacts were reviewed by a single reviewer and, for those studies meeting the eligibility criteria, full-text articles were obtained. In total, 542 articles were included for review.
English-language articles and health technology assessments.
RCTs and observational studies of OCT and AMD or DME.
Studies focusing on either diagnosis or monitoring of disease.
Studies in which outcomes were not specific to those of interest in this report.
Studies of pediatric populations.
Studies on OCT as a screening tool.
Studies that did not assess comparative effectiveness of OCT with a referent, as specified below in “Comparisons of Interest”.
Outcomes of Interest
Studies of sensitivity, specificity.
Comparisons of Interest
Evidence exists for the following comparisons of interest:
OCT compared with the reference “fluorescein angiography” for AMD.
OCT compared with the reference “biomicroscopy” or “stereo or fundus photography” for DME.
Summary of Existing Evidence
No evidence for the accuracy of SD OCT compared to either FA, BM or SFP was published between January 2006 to February 2009; however, two technology assessments were found, one from Alberta and the other from Germany, both of which contain evidence for TD OCT. Although these HTAs included eight studies each, only one study from each report was specific to this review. Additionally, one systematic review was identified for OCT and DME. It is these three articles, all pertaining to time and not spectral domain OCT, as well as comments from experts in the field of OCT and retinal disease, that comprise the evidence contained in this review.
Upon further assessment and consultations with experts in the methodology of clinical test evaluation, it was concluded that these comparators could not be used as references in the evaluation of OCT. The main conclusion was that, without a third test as an arbiter, it is not possible to directly compare the sensitivity and specificity of OCT relative to either FA for AMD and stereo- or fundus – photography for DME. Therefore, in the absence of published evidence, it was deemed appropriate to consult a panel of experts for their views and opinions on the validity of OCT and its utility in clinical settings. This panel consisted of four clinicians with expertise in AMD and/or DME and OCT, as well as a medical biophysicist with scientific expertise in ocular technologies. This is considered level 5 evidence, but in the absence of an appropriate comparator for further evaluation of OCT, this may be the highest level of evidence possible.
Summary of Findings
The conclusions for SD OCT based on Level 5 evidence, or expert consultation, are as follows:
OCT is considered an essential part of the diagnosis and follow-up of patients with DME and AMD.
OCT is adjunctive to FA for both AMD and DME but should decrease utilization of FA as a monitoring modality.
OCT will result in a decline in the use of BM in the monitoring of patients with DME, given its increased accuracy and consistency.
OCT is diffusing rapidly and the technology is changing. Since FA is still considered pivotal in the diagnosis and treatment of AMD and DME, and there is no common outcome against which to compare these technologies, it is unlikely that RCT evidence of efficacy for OCT will ever be forthcoming.
In addition to the accuracy of OCT in the detection of disease, assessment of the clinical utility of this technology included a rapid review of treatment effects for AMD and DME. The treatment of choice for AMD is Lucentis®, with or without Avastin® and photodynamic therapy. For DME the treatment of choice is laser photocoagulation, which may be replaced with Lucentis® injections (Expert consultation). The evidence, as presented in systematic reviews and other health technology assessments, indicates that there are effective treatments available for both AMD and DME.
Considerations for the Ontario Health System
OCT testing is presently an uninsured service in Ontario with patients paying approximately $150 out-of-pocket per test. Several provinces do provide funding for this procedure, including British Columbia, Alberta, Saskatchewan, Newfoundland, Nova Scotia, Prince Edward Island, and the Yukon Territory. Provinces that do not provide such funding are Quebec, Manitoba and New Brunswick.
The demand for OCT is expected to increase with aging of the population.