Optic disc edema is a critically important finding on physical examination. It warrants immediate attention by the clinician and often prompts extensive patient evaluation. But for both physicians and researchers who deal with both the clinical and pathophysiologic aspects of optic disc swelling, they must “speak the same language.” Over 3 decades ago, Lars Frisén, MD, a Swedish neuro-ophthalmologist, devised and published a grading system for optic disc edema. In particular, the system has been applied to the description of papilledema. The term “Frisén grade” is now embedded in the medical literature.
In this following article, Lars Frisén shares his experience in developing his grading system of optic disc edema. To be sure, we owe Lars Frisén a great debt of gratitude.
In the late 1960s, the daily routines of this neuro-ophthalmologist-wannabe at the Sahlgren University Hospital in Gothenburg, Sweden, then one of Europe's largest, included running around on the neurology and neurosurgery wards to help monitor patients with high intracranial pressure. This was long before anyone had heard of bedside fundus photography, much less computed imaging, or continuous intracranial direct pressure measurements. Many of my neurology and neurosurgery colleagues were quite accomplished ophthalmoscopists, but they also were well aware of the wide variations between examiners. Some categorized optic disc swelling in vague terms such as “incipient,” “moderate,” “florid,” “choked” or “vintage.” Others concentrated on protrusion, which was roughly and highly variably estimated in terms of dioptric elevations. Some paid much attention to venous pulsations, whereas others relied on the prominence of small vessels at the disc border. With my ophthalmology background, I preferred the arcane technique of monocular indirect ophthalmoscopy, which offered a better overview than the direct counterpart, even with small pupils. Still, terminology remained a major obstacle to reproducible evaluations. The first hints of better prospects came from a somewhat unexpected direction, namely, novel funduscopic observations of optic atrophy. Using a green filter to produce so-called red-free illumination, William F Hoyt, MD, at the University of California at San Francisco had found a way to visualize the peripapillary part of the retinal nerve fiber layer and the changes that occur with optic atrophy.
A GREEN FUTURE
The era of “the greening of the fundus,” so named by Joel Glaser, MD, a former fellow of Dr Hoyt, began around 1970. I was very fortunate to obtain a 1-year fellowship with Dr Hoyt in 1972, on the 7th floor of the San Francisco Medical Center, with a gorgeous view over the Bay Bridge. In the opposite direction, on Dr Hoyt's desk, there was another gorgeous view, namely, heaps of absolutely stunning black-and-white fundus photographs, mostly from patients with different stages of glaucoma. These photographs had been produced by a highly skilled fundus photographer, Ronald Eckelhoff, by means of red-free illumination, superhigh magnification, black-and-white film, and high-contrast processing. Side-by-side with some of the photographs were beautiful graphic renditions made by the medical illustrator Joan Esperson Weddell. Nancy M Newman, MD, also made important contributions.
The image sets were repeatedly presented to the fellows and to the constant stream of domestic and foreign visitors. While everyone was highly intrigued, everyone also was most cautious of Dr Hoyt's descriptions and interpretations. Retrospectively, this was in part attributable to the then lack of coherent and concise terminology. Many were highly experienced ophthalmoscopists, and they had certainly never seen anything like the curious light and dark lines and wedges that coursed along the arcuate bundles in the photographs. Many admitted to never having seen a retinal nerve fiber layer at all.
Having worked as a part-time junior teacher (“amanuensis”) in anatomy during my medical student years, I had some knowledge of anatomic terminology. To some degree, this background helped to chisel-out Dr Hoyt's concepts of nerve fiber layer atrophy. Discussions were sometimes a bit tense, like “Lars, can you see the difference between a horse and a cow?”, but my Swenglish idiom often seemed to have a mollifying effect. Incidentally, Dr Hoyt liked to introduce me to visitors as a fellow from polar bear country (but there are no polar bears in Sweden).
I spent much time on my own in the hospital's well-equipped photographic darkroom, trying to enhance image quality, particularly by means of unsharp masking. This was very tedious work, often involving exposure times of the order of 20–30 minutes. This was a far cry from modern image processing, where unsharp masking is just a mouse-click away. Incidentally, manual and computed unsharp masking differs in one important aspect; the former levels out brightness differences across the full image which the latter cannot do. Nobody had heard of personal computers in those days. Apropos fine image detail, it should be mentioned that only a handful of scientific journals had paper quality good enough to satisfy Dr Hoyt.
Together with Carol L Knight, MD, Dr Hoyt had made some seminal observations on nerve fiber layer changes in papilledema, but there were few opportunities to pursue this subject during my stay at UCSF. However, an increasing number of reports from other sources pointed to a possible role of axoplasmic flow, or more precisely interference with axoplasmic flow as a major factor in the generation of disc swelling. The reason for the block of axonal transport with raised intracranial pressure was (and is) not well understood. The obstructed flow must be quite marginal to explain why papilledema typically develops slowly, over days or weeks. In contrast, acute ischemia may cause axonal swelling within minutes.
IN THE BACK OFFICE
Back in Gothenburg, renewed attempts toward photographic enhancements and often repeated scrutiny of my large collection of disc swelling photographs ultimately led to a model of stages of swelling that was based solely on features attributable to swollen axons. The reasoning was as follows.
Blocking of axonal flow causes an accumulation of axonal constituents at the site of the disturbance. The accumulation leads to localized distension and elongation of the affected axonal segments, resulting in increased reflectivity, decreased transparency, and a break up of the fine radiating bundle detail that is just at the limit of both photographic and ophthalmoscopic resolution.
The first sign of increased reflectivity of swollen axons is the appearance of a faint gray cast over and just outside the disc border. In early stages, this cast may be best appreciated in the wide-field view provided by indirect ophthalmoscopy. Decreased transparency is harder to see in early stages, but it causes a desaturation of the background color, it obscures detail within the nerve fiber layer itself, for example, minor branches of the retinal vessels, and it obscures pigmentation detail in underlying layers.
These early changes are not uniformly distributed across and around the optic nerve head. This is due to regional variations in the diameters and number of axons that traverse the scleral opening. The prominence of swelling varies roughly in proportion to the number and diameter of axons at a particular location. Hence, swelling first appears in the superior and inferior poles, followed by the nasal disc sector, and finally the temporal sector. It is unfortunate that the pole areas are difficult to evaluate for early changes because of the normal crowding of axons and the passage of major retinal vessels. The next best place to look for minimal disc swelling is the nasal sector. If the nasal sector is normal, it is safe to predict that the temporal sector also will be normal.
With increasing swelling, the scleral opening becomes obscured, the apparent disc diameter increases, the optic cup begins to fill in, major vessels become partly enveloped by swollen axons, and venous outflow becomes compromised. Ultimately, the nerve head will assume a dome shape with a completely filled-in cup and a steep slope to the surrounding retina.
THE STAGING SCHEME
The normal regional variation in distribution of axons and the predictable sequence of events form a simple scheme for clinical grading of optic disc swelling (Table 1). The various stages are intentionally defined without overlap to allow a single digit to describe the amount of optic disc swelling in any given patient. Transitional staging is discouraged.
Note that the scheme deliberately excludes secondary effects of swelling like hyperemia, hemorrhages, localized infarcts, and micro drusen. Such changes must not influence on the evaluation of the swelling itself, but they also need to be recorded. Further, there must be no signs of optic atrophy.
Photographic illustrations of the various stages were included in the original 1982 article (1), which can be accessed at http://jnnp.bmj.com/content/45/1/13.long. Unfortunately, these photographs do not reproduce very well. Better examples can be viewed at http://www.oft.gu.se/webdiagnos/ (select Nerve Fiber Layer Basics, then select Swelling). Unfortunately, fundus cameras have difficulties capturing fine detail in higher grades of swelling because of their small depth of focus. However, fundus photographs do accurately portray the large range of normal variations of background colors and patterns, variations that tend to work against generalizations. Therefore, there seems to be a good rationale for artistic drawings (Fig. 1).
Stage-1 optic disc edema is the one that is most difficult to recognize, particularly in discs of larger than average diameter, where axons are less crowded at the disc border than in average-size discs. Conversely, discs with smaller than average diameters have more crowded axons. These normal variants lack the key features of swollen axons, that is, increased opacity, reduced transparency, and contorted bundle detail. Serial fundus photographs may help to identify minimal ongoing changes. Examination of the narrow parapapillary ring reflex and minute punctate highlights (“Gunn's dots”) may also be helpful. The latter tend to disappear with swelling of axons, presumably because of mechanical deformation of the funnels of the glial cells that traverse the nerve fiber layer. The ring reflex often becomes fainter, deformed, and fragmented, presumably as reflection of minute irregularities in the normally smooth surface of the nerve fiber layer. The ring reflex retains its normal appearance in conditions mimicking papilledema, where there is no swelling of axons. The ring reflex is best evaluated with indirect ophthalmoscopy.
The staging scheme is not restricted to optic disc swelling from high intracranial pressure but is equally applicable to other causes of disturbed axonal flow, including ischemic and inflammatory optic neuropathies. The sole provision is that there must be no signs of axonal wasting; any atrophy upsets the normal sequence of events.
Seen through the retrospectoscope, the anatomic reasoning behind the staging scheme still seems valid. However, seen through the clinician's eye, the scheme would have benefitted from the inclusion of an additional stage, in between Stages 2 and 3. Regrettably, I have not been able to find a solution. The scheme has another possibly uncomfortable feature in that it defines an ordinal scale with unequal steps. Hence, statistical analyses should never use mean values and SDs.
The illustrations were skillfully produced by David Fisher at the Medical Education and Design Services, Birmingham, Alabama.