Early detection + Progression analysis

Glaucoma vacillates its course; clinicians can depend on OCT for this technology’s objectivity.

By Kelly Lee, BA and Albert S. Khouri, MD

Even with the variety of tools and tests available (tonometry, visual field [VF] testing), evaluating a glaucoma diagnosis and disease progression remains challenging, partly because glaucoma has a variable course and is a slowly progressing disease that can lead to blindness.

Glaucoma treatment can be started soon after diagnosis to slow its advancement and preserve vision; but how do we determine who should be observed more carefully? With OCT, this question becomes less ambiguous. This technology can detect wedge nerve fiber layer defects and macular thickness changes earlier than a clinical exam or prior technologies.

OCT looks at the retinal layers via low-coherence interferometry to create a reproducible, high-resolution cross section of the layers. The images compose data sets that represent the backscattering of light. OCT has evolved beyond the improved image resolution and speed of spectral domain OCT (SD-OCT), the standard model used today. Swept-source OCT is one of the latest additions recently approved in the US with speeds twice as fast as SD-OCT and a deeper range of imaging.¹ However SD-OCT remains the predominant model used today and will be the focus here. As the technology advances, studies show the newer OCT models to have equal or greater efficacy in glaucoma diagnosis.²

The problem with its predecessors

Prior to OCT, some available glaucoma tests yielded subjective and often irreproducible results (as in the wide VF variability noted in the Ocular Hypertensive Treatment Study) that focused on functional changes.³ For example, VF tests would sometimes detect changes only after the disease advanced in course with significant nerve fiber layer loss, emphasizing that structural progression may precede functional loss in the majority of patients. Disease diagnosis could be delayed by years because of this incongruence.

The clinical exam previously involved serial photographs of the optic disc, which proved to have its own set of challenges (the need for an experienced photographer, difficulty detecting subtle changes in nonstereoscopic images, etc.).

But now clinicians can rely on OCT, which has defined its role in the early detection of glaucoma and its progression. With more objective evidence obtained through OCT and the correlation of these OCT-obtained abnormalities with clinical findings and functional tests, clinicians can identify those patients who should receive treatment or possibly interventional surgery in a timely fashion.

Glaucoma detection

To detect the disease early, we can use OCT to look for glaucomatous changes in multiple parameters, such as the retinal nerve fiber layer (RNFL), optic nerve head (ONH) and the macula. The RNFL thickness is commonly looked at first and referenced to a normative database (more on this below). There are many mechanisms for glaucoma, but the ultimate final pathway is death of the RNFL (manifested as thinning). OCT can also calculate optic disc areas and cup-to-disc ratios using radial line scans. However, due to the wide variations in optic nerve anatomy (size and morphology), this remains a less useful test for glaucoma.

OCT can help to define the paramacular region as an area of interest.⁴ Loss of retinal thickness in this area manifests in glaucoma patients early in the disease process, and this loss continues as the disease progresses. To analyze disease progression over the duration of follow-up (every three to six months depending on risk factors, IOP and clinical picture), clinicians need to have images taken — at baseline and every one to two years after — that are accurately aligned (image registration). Inherent variability in measurements and artifacts must be low to avoid noise and error. The physician must also take age and race into account when looking at the progression of changes in retinal layer thickness (we will discuss these sources of limitations below).

Progression analysis

After collecting the data, physicians must determine the significance of any changes and whether they are disease- or normal, age-related occurrences. Glaucoma progression algorithms come in two flavors: trend-based and event-based approaches.

Event-based analysis looks at how a measurement deviates from its baseline; trend-based analysis monitors changes over time using regression analysis. Most software shows trend and event analyses to detect RNFL progression by visually depicting the areas of significant change by color. The measurements are also classified by quadrants and smaller sectors, which can show focal changes that might otherwise be buried in a single average value. Studies have examined patterns of RNFL progression, such as widening of an existing RNFL defect, deepening without widening of an existing RNFL defect, or development of a new defect (See Table 1).

The inevitable limitations

One challenge with using OCT in glaucoma detection is that commonly used devices use different software algorithms and their own age-matched normative database to compare measurements. The databases vary in size (from about 200-500 patients of all ages), are heterogeneous and differ in the included racial groups.

Each database also has its own exclusion and inclusion criteria based on ocular and medical conditions, with no uniformity among platforms. This brings into question how generalized this normative database is. One way to overcome this in clinical practice is to use the patient as his own normative database, comparing measurements to previous exams.

Another challenge to examining these changes is determining if they are statistically relevant (Are they due to disease progression, age-related changes or short-term measurement fluctuations?) This discrimination is difficult but may be resolved with better normative databases and improvements in OCT hardware and software.

Clinical applications

OCT software determines RNFL thickness by taking an average value of multiple scans. Spatial averaging can take the average of differently sized regions so that smaller areas will exclude extreme values. Each OCT device has its own degree of spatial averaging to eliminate “image noise,” which is also why RNFL measurements cannot compare across platforms. The RNFL measurement may be only slightly out of the range of normal values. As such, clinicians should avoid relying on color-coded printouts to make their decisions.

Standard OCT reports depict normal RNFL thickness as green, while yellow is borderline and red signifies abnormal values. The OCT reading sometimes overlooks subtle abnormalities as long as they fall within the reference database range and state that the patient has normal findings (see Figure 1).

Figure 1. A clinical example of OCT missing an abnormality found in a glaucoma patient. A) Right ONH splinter hemorrhage superior-temporal, which is an indicator of progression in glaucomatous eyes and has been linked to progressive retinal nerve fiber layer thinning.⁵ B) OCT reflectance image of the optic disc, which also shows the subtle hemorrhage. C) OCT color-coded quadrant showing that the patient’s RNFL is within normal limits.

Although OCT is useful in early to moderate glaucoma, it is not as helpful in advanced stages due to the “floor effect” on RNFL thickness. So, with advanced RNFL loss the OCT cannot measure further loss in thickness, and the measurements stabilize at “floor levels” as the OCT measures thickness of remnant tissues including glial cells that partly compose this layer. For this reason, central VF testing becomes more essential in monitoring and obtaining a global picture of the progression.

These points are memorable:

• Only look at the best quality scans

• Identify artifacts and limitations in measurements in certain eyes due to myopic changes or tilted discs.

• Dry eye disease, cataracts and other media opacities could impact OCT.

• Low signal strength due to the above factors can result in deceivingly low RNFL values, therefore overestimating the state of disease or its progression.

Summary

OCT is objective, quantitative and continues to evolve in reproducibility and resolution. OCT complements other diagnostic tools; it is not a substitute. Current OCT technologies detect structural changes that in the past would have been unappreciable, resulting in delayed detection of disease. Along with the parameters discussed, OCT technology may become even better in detecting glaucoma and its progression. For example, new-generation OCT angiography can shed light on possible retinal vascular changes and further our understanding of the disease.

Glaucoma diagnosis is not a black or white, structure-versus-function battle. A clinician must take every piece of information available to paint a larger picture and eliminate the guessing game. For this reason, OCT has since become a staple in virtually every ophthalmologist’s office, lending an objective eye to our more subjective one. With early detection becoming the key to decreasing the risk of potential blindness, OCT has become a critical component in our standard of care for glaucoma patients. OM

REFERENCES

About the Authors
	Kelly Lee is a third-year medical student at Rutgers New Jersey Medical School. She reports no financial disclosures.
	Albert S Khouri, MD is assistant professor, associate director of glaucoma, and program director of the ophthalmology residency at Rutgers New Jersey Medical School. Disclosures: speaker bureau for Alcon, Allergan; consultant for Topcon; grant support from Allergan, NJ Health Foundation, Fund for NJ Blind.