J Neuroophthalmol 2001 December;21(4):302-309
Nicholas J. Volpe, MD
Departments of Ophthalmology and Neurology, Scheie Eye Institute, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania.
JOURNAL OF NEURO-OPHTHALMOLOGY 2001;21:302-309
Optic nerve disorders were not reliably diagnosed until the late nineteenth century when ophthalmoscopy became part of the ophthalmic examination. By the early 1900's, all of the salient clinical features of optic neuritis and its relationship to “systemic sclerosis” were recognized, but there was much controversy and misunderstanding about its differential diagnosis, pathogenesis, and possible treatment. During the twentieth century, physicians began to distinguish optic neuritis from infectious, hereditary, toxic, nutritional, and ischemic optic neuropathies. The development of magnetic resonance imaging and the results from recent clinical trials have enhanced our understanding of the relationship between optic neuritis and multiple sclerosis. The next decade holds the promise of further elaborating the pathogenesis and treatment of optic neuritis.
Much of the clinical profile of optic neuritis had been established by the late 1800's. The invention of the ophthalmoscope by Helmoholtz in 1845 allowed ophthalmologists to differentiate types of “amaurosis” and to localize vision loss to the optic nerve. By the 1880's, von Graefe (1) and Nettleship (2), had described many of the salient features of the clinical syndrome that we now call idiopathic optic neuritis. Several individuals were accumulating large series of patients by the turn of the century. In 1884, based on a series of 28 patients, Nettleship offered the following description of patients with optic neuritis:
The subsequent literature abounds with discussions of infectious or toxic etiologies, and proposals of barbaric treatments often claimed to be successful because of the under-appreciated rate of spontaneous recovery. The distinction between optic disc swelling secondary to papilledema and other causes of optic disc swelling was reported in the early part of the twentieth century. The association of optic neuritis with multiple sclerosis was also established at that time. However, except for increasing clinical sophistication allowing better distinction of optic neuritis from other forms of disc swelling and optic neuropathy, very little further insight had been gained into this disorder until 15 years ago. Patients could be counseled in 1985 much the same way they could be counseled in 1885. (The text Optic Neuritis and Its Differential Diagnosis by Perkin and Rose (3) published in 1979 summarizes the available literature until that time.) It was not until 140 years after the invention of the ophthalmoscope, with the application of the randomized clinical trial, magnetic resonance imaging, and the science of immunomodulation to the study of optic neuritis, that we have been able to make dramatic strides into understanding the etiology, treatment, and prognosis of idiopathic optic neuritis.
OPTIC NEURITIS IN ANTIQUITY
Before the invention of the ophthalmoscope in 1845, diseases of the fundus were impossible to recognize. Ancient physicians divided all eye diseases into “ophthalmia” or “blindness”. “Ophthalmia” included all varieties of conjunctival and corneal diseases recognized through inspection of the globe. “Blindness” referred to vision loss not based on an obvious change in the visible surface or media of the eye. Blindness was often considered a divine punishment for sin. Within that group, optic neuritis may have accounted for some of the “miraculous” spontaneous cures.
The earliest references to optic nerve dysfunction as a mechanism for vision loss are found in Arabic texts of ophthalmology written in the ninth century. In what is believed by some to be the first major textbook of ophthalmology, Hunain Ibn Is-Haq (4) described three different forms of paralysis of the eye: those involving perception alone, those involving eye motion alone, and those involving both (4). In these early descriptions, he does not distinguish optic neuritis from other diseases of the posterior segment. However, there are specific references in this volume (and others) to pain and heaviness, “swelling” of the optic nerve, and the afferent pupillary defect (4):
Another Arabic writer, Ali Ibn Isa, referred to inflammation of the optic nerve, although he may have been describing papilledema, as he related “the cause of the blindness to the ventricles of the brain.” (5) In his textbook of ophthalmology, Isa describes various affections of the optic nerve, including those resulting from “warmth, cold, humidity etc”.
OPTIC NEURITIS IN THE PRE-OPHTHALMOSCOPE ERA
In one of the earliest English language textbooks of ophthalmology, published in 1823, George Frick (6) wrote that optic neuritis is:
His description suggests that, without the advantage of the ophthalmoscope, it was impossible to distinguish uveitis, migraine, retinal detachment, optic nerve and orbital apex disorders. Frick makes reference to severe pain in the orbit preceding vision loss, abnormal responses of the pupil to light, and the visual behavior of patients with central scotomas:
Manuals written by Saunders (7) in 1821 and Littell (8) in 1846 present a similarly confused impression of optic nerve and retinal diseases.
OPTIC NEURITIS IN THE POST-OPHTHALMOSCOPE ERA
With the widespread use of the ophthalmoscope in the second half of the nineteenth century, the various diseases affecting the optic nerve and retina became distinguishable. Optic neuritis was first described with some uniformity by von Graefe in 1860 (1) and by Nettleship in 1884 (2). Nettleship acknowledged that Leber, Hutchinson, and Hock had previously described cases of optic neuritis. Nettleship points out confusion with tobacco amblyopia in Leber's writing and describes Hutchinson's series as having a variable group of patients and Hock as describing the characteristic pain of optic neuritis (which Nettleship attributed to “stretching of the inflamed optic nerve sheath.”) Nettleship states that:
Although he identified all the salient features of optic neuritis, he did not mention a relationship to other relapsing and remitting neurologic symptoms. Nettleship commented on the failure of sight being noticed quickly and reaching its worst within two to three days. He reported that pain usually began shortly before vision loss and that it worsened with eye movements or pressure on the globe. Vision loss is vividly described:
Eleven of his sixteen patients displayed a central scotoma, and three also demonstrated peripheral field constriction. He described patients with healthy appearing optic nerves and others where there were “decided, if slight, changes much earlier.” He suggested that the cases with normal optic nerves might have had a periostitis in the optic canal, but among those with disc swelling, he could not “suppose the mischief to be seated so far back”. He noticed the most significant recovery to be complete within 4 to 6 weeks. Up to nine of his patients were thought to have syphilis, although he did not specifically implicate this infection as the cause of vision loss. He distinguished “neuritis” patients with optic nerve head swelling from those with increased intracranial pressure resulting from anterior cerebral lobe tumors (papilledema) by the presence in the latter of vomiting, convulsions, and other cerebral symptoms.
Parinaud (9), Uhthoff (10), Buzzard (11), and Gunn also contributed early descriptions of optic neuritis (12). Parinaud described dyschromatopsia. Uhthoff detailed the visual field defects, various patterns of optic disc pallor, and the transient blurring of vision occurring with exercise that has become to be known as “Uhthoff's symptom.” Buzzard's 1893 paper summarizes these contributions and contains a detailed discussion of optic disc pallor as part of multiple (“disseminated”) sclerosis. Among 100 patients with disseminated sclerosis, he found 42 with optic disc pallor, and emphasized the importance of vision loss with disc pallor in distinguishing true from “hysterical” forms of disseminated sclerosis. Buzzard credits Charcot (13) with the first description of “amblyopia” as a frequent symptom of cerebrospinal disseminated sclerosis. Buzzard writes:
Buzzard distinguished between the disc pallor of disseminated sclerosis and that of tabes dorsalis. Tabes, he believed, had more severe and permanent vision loss and more marked disc pallor:
Based on his experience with approximately 350 cases of optic neuritis, Gunn (3,12,14,15) concluded that an infectious etiology was likely. He divided cases into isolated “acute retro-ocular neuritis” without disc pallor or other neurologic manifestations and those in which there were acute episodes of optic neuritis as well as disc pallor and other neurologic manifestations. He described the presence of visual loss as a feature that distinguished optic neuritis from conditions in which disc swelling was associated with intracranial disease such as tumor, wherein chronic disc swelling was associated with preserved visual function (12). He reasoned that central scotomas were common in optic neuritis because the “optico-macular fibers are the most active physiologically in function and therefore will undergo most rapidly the normal degenerative changes.” He disagreed with contemporary authors who argued that the direction of eye movement associated with the greatest periocular pain could be used to predict the nature of the field defect.
Gunn divided “retro-ocular neuritis” into three groups, those associated with: 1) infection of the orbit, paranasal sinuses or meninges 2) systemic infectious diseases (syphilis, tuberculosis), and 3) disseminated sclerosis. He postulated that disc pallor is not an indication of atrophy of the nerve fibers but of an increase in connective tissue. He also described a relatively unsustained pupillary constriction to direct light in the eye affected with optic neuritis, a contribution that led to his receiving (undeserved) credit for the “Marcus Gunn pupil:”
SYMPTOMS AND SIGNS OF OPTIC NEURITIS
The periocular pain characteristic of optic neuritis has long been recognized. Nettleship (2) and Swanzy (16) thought that more severe pain was associated with a worse visual prognosis. Hock postulated that the pain was the result of inflammation of the optic nerve sheath (17). (The prevalence of pain in the large reported series of optic neuritis is summarized in Table 1. Perkin and Rose (3) reported that periocular pain persisted longer than four weeks in 25% of patients.
TABLE 1. Reported prevalence of pain in patients with optic neuritis
Most authors have found central scotomas to be the most common pattern of field loss, followed by peripheral field constriction (2,3,11,12). Adie (18) and Traquair (19) argued that involvement of the central field was a prerequisite for diagnosis. Most series have emphasized central visual loss, with a small percentage of patients having altitudinal or hemianopic vision loss (3,20). In a 1953 analysis of 100 consecutive optic neuritis cases, Chamlin (21) found that 25% of patients had non-central nerve fiber bundle defects. Positive visual phenomena were first reported by Gunn (12) and Traquair (19). Momentary sparkles—spontaneous or evoked by sound or eye movement– were not described until the late twentieth century (22,23). Uhthoff (10), and later Gunn (12,15), described transient blurring of vision during exercise. Subsequently, Uhthoff's symptom has been recognized to be evoked by increased body temperature, and has been described in other optic neuropathies. Later studies (24–26) reported that patients with Uhthoff's symptom have an increased likelihood of MRI abnormalities and risk of developing multiple sclerosis.
From the earliest descriptions of optic neuritis, the high prevalence of normal-appearing optic nerves at symptom onset has been well recognized. This diagnostic feature led to ready distinction of optic neuritis from other diseases of the optic nerve and retina. Uhthoff described disc swelling in only 5% of his optic neuritis patients (27). (The prevalence of disc swelling in large series reports is summarized in Table 2.) The prevalence of optic disc hemorrhages has been low (28). Lillie (29) in 1934 and Berliner (30) in 1935 described patients with retinal hemorrhages only with marked disc swelling.
TABLE 2. Reported prevalence of disc swelling in patients with optic neuritis
The most significant diagnostic dilemma in the era before neuroimaging was the clinical distinction of optic neuritis from vision loss secondary to compressive optic neuropathy and papilledema. In Jackson's 1900 textbook (30), these entities are not clearly distinguished. Paton (31,32) emphasized that in some cases of optic neuritis, the disc swelling can be as severe as that of papilledema. In 1922, Dandy (33) wrote that “intracranial tumors have too infrequently been missed by the over zealous enthusiasts searching for optic neuritis.” Together with Traquair (19) and Lillie (29), Dandy emphasized that the rapidity of vision loss and the type of visual field loss are the most reliable signs to distinguish these disorders. In general, papilledema has been easily distinguished from optic neuritis based on the relative preservation of vision, the associated symptoms (nausea and seizures), and the degree and persistence of the disc swelling (2,12,15,27). However, Walsh and Ford (34,35) emphasized that central scotomas, relatively rapid onset of vision loss, and spontaneous improvement could occur with compressive optic neuropathy.
Cerebrospinal fluid analysis
The importance of CSF analysis in optic neuritis was not recognized until the past half century. In 1939, Watkins (36) reported CSF findings in 40 patients with optic neuritis but did not tie the findings to multiple sclerosis. In subsequent series of optic neuritis patients, elevated levels of gamma globulin in the CSF were found in about 10–24% of patients and increased cell counts were found in 50% of patients (37,38). One of the six patients in the series of Sandberg and Bynke (37) demonstrated oligoclonal banding. Perkin and Rose found abnormal cell counts in 15% of their patients and elevated protein in 12% (3). The Optic Neuritis Treatment Trial (ONTT) showed that CSF analysis is a much less robust predictor than MRI of the future development of multiple sclerosis (39).
ETIOLOGY OF AND DIFFERENTIAL DIAGNOSIS OF OPTIC NEURITIS
The term “optic or retroocular neuritis” has had many different meanings. In the 1946 (first) edition of Clinical Neuro-ophthalmology by Walsh (41), “optic neuritis” includes hereditary, toxic, cancer-associated, infectious and inflammatory etiologies. In the 1980 edition Neurology of the Visual System, Cogan (40) includes inflammatory and vascular etiologies under “optic neuritis.” In the post-ophthalmoscope era, optic neuritis has been attributed to infections, hereditary disorders, sinus disease, compression and toxins. Before that era, it was attributed to “colds”, gout, rheumatic disease and tobacco (2,11,16,19,27). Many authors mistook the high rate of spontaneous resolution to reflect “successful treatments” such as sinus surgery or the elimination of a toxin. Here are some of the more prevalent ideas:
Sinus and tooth disease
The proximity of the sinuses to the optic nerves naturally led to the positing of a causative relationship (42,43). In 1897 Gunn wrote,
This was further supported by operative and autopsy evidence of active sinusitis extending through bony perforations to reach the optic nerve and cause focal inflammatory lesions. In 1933, Benedict (44) found that only 1 of 233 cases of retrobulbar neuritis examined at the Mayo clinic was attributable to sinus disease. In the initial review of the Johns Hopkins experience in 1931, Woods and Roland (45) found that 8% of patients had optic neuropathies resulting from sinus disease. However, in a 1952 review from the same institution (133 new cases), Bagley (46) did not find a single case that was attributable to sinus disease. Others believed that optic neuritis resulted from infections of the teeth and tonsils (47). Yet Traquair was unable to find a single case that resulted from tooth abscess or to have a “rhinogenic” etiology (19). In Gifford's 1931 review of 203 reported cases (48), he found only three cases convincingly caused by sinus disease. Nettleship (2), Percival (49), Gunn (12), and Swanzy (16) believed that inflammatory compression of the optic nerve within the optic canal was the cause of optic neuritis. This theory led to widespread use optic canal decompression as a method of treatment.
Many cases of optic neuritis in the past were undoubtedly the result of syphilis (2). Buzzard believed that syphilitic optic neuropathy differed from optic neuritis in having a rapid progression to complete blindness (11). In the series of optic neuropathy patients from Johns Hopkins reported in 1931, 23 were attributed to syphilis (45). Those attributed to syphilis were almost always bilateral and associated with disc swelling. Walsh (50) concluded in 1956 that inflammatory optic neuritis was rare in acute syphilis, and that optic nerve involvement was a more prominent feature of the latter stages of the disease associated with more generalized neurologic disturbance. Nearly every other known infectious agent (bacterial, fungal and parasitic) has been reported to cause optic neuropathy but the presentation is ultimately readily distinguished from idiopathic optic neuritis (except for cat scratch disease from Bartonella infection).
Ischemic optic neuropathy
The early papers on optic neuritis include cases that would today be considered to represent non-arteritic anterior ischemic optic neuropathy (NAION):
Even with current sophisticated diagnostic testing, ischemic optic neuropathy and optic neuritis may be hard to distinguish (51). Most of the literature concerning AION is written after 1970, indicating that it was either less common or less well recognized in earlier times. Given that many patients considered in past days to have had “optic neuritis” did not recover vision and displayed substantial disc swelling, it is likely that AION was mistakenly included in older series of “optic neuritis.”
Hereditary, toxic, and nutritional optic neuropathies
With the exception of Leber's hereditary optic neuropathy (LHON), hereditary optic neuropathies were rarely confused with optic neuritis. However, patients with optic neuritis and LHON both commonly present with the sudden onset of vision loss and central scotoma. Considering that at least 50% of patients with LHON lack a family history of the disease, it seems likely that LHON was overlooked in earlier series of patients reported to have optic neuritis.
Toxic and inflammatory optic neuropathies were often confused in the past. Treating physicians withdrew the offending agent, patients improved, and cause and effect were presumed. Since Traquair's original report in 1930 (19), in which he identified 1088 of tobacco-amblyopia cases from the Royal Infirmary in Edinburgh, this entity has been confused with optic neuritis. “Tobacco amblyopia” represented 6% of 506 cases of optic neuropathy from a Johns Hopkins series reported in 1952 (46).
RELATIONSHIP OF OPTIC NEURITIS TO MULTIPLE SCLEROSIS
As long as optic neuritis has been recognized, its relationship to multiple sclerosis has been appreciated. In 1893, Buzzard reported five patients with a history of disseminated or multiple sclerosis that had episodes of visual failure with recovery consistent with optic neuritis (11). In the early portions of the twentieth century, there was controversy over the relationship of optic neuritis to multiple sclerosis. In 1930, Adie asserted that multiple sclerosis accounted for all cases of optic neuritis:
Adie pointed out that optic neuritis might “be the only manifestation in a lifetime of the activity of the agent causing multiple sclerosis.” (52) Percival (49) never saw a single case in which optic neuritis progressed to multiple sclerosis. Yet in 1934, Lillie (29) found that 75 (15%) of 500 cases of multiple sclerosis had optic neuritis as their initial event and nearly 200 developed it later in the course of their illness.
Over the first 80 years of the twentieth century, 30 studies reported the rate of the development of multiple sclerosis after an episode of optic neuritis (3). Perkin and Rose (3) emphasized the difficulty in interpreting this literature because of the inexact inclusion criteria, questionable neurologic evaluations at presentation, variable length of follow up, and criteria used to establish a diagnosis of multiple sclerosis. The rate at which multiple sclerosis developed has varied between 12% and 85% in these series. Inadequate diagnostic criteria and clinical information prohibit valid comparisons of most of these retrospective studies. Recent prospective studies have yielded more valid, but not necessarily less confusing, information (summarized in Table 3).
TABLE 3. Summary of prospective studies reporting the proportion of optic neuritis patients developing multiple sclerosis
TREATMENT OF OPTIC NEURITIS
The history of the treatment of optic neuritis centers on the effort over the past 50 years to determine whether systemic corticosteroids are valuable. Until the latter part of the nineteenth century, however, treatment revolved around the placement of solutions in the eye (“collyria”). Various compounds, including spittle, were rubbed in the eye (53). In the ninth century, Is-Haq wrote,
In the 1730's, Chevalier Taylor (54) suggested that “rubbing the lower part of the globe with a small instrument frees up the nerves destined for the motion of the pupil.” He also pricked the muscles of the eye with a blunt needle and bled patients through the jugular vein. Frick's textbook in 1823 (6) suggests the use of emetics for treatment of the various forms of “amaurosis”.
Once the specific syndrome of optic neuritis was recognized over 100 years ago, treatment centered around emetics, identifying toxic substances, sinus surgery, and optic canal decompression. In 1899, Fuchs (55) suggested that the “treatment of the disease is that of neuritis in general; in the acute stage energetic diaphoresis proves particularly efficient.” In the early part of the twentieth century, many patients were treated with sinus surgery or optic canal decompression. Optic canal decompression was still suggested for the treatment of retrobulbar neuritis as recently as 1952 (56). In 1933, Benedict (44) recommended treatment with intravenous typhoid vaccine, postulating that this increases the vascularity of the tissues (the same mechanism proposed to explain the apparent benefit gained from operating on non infected sinuses). In 1941 Duke-Elder (57) directed treatment at the reduction of local edema and advocated the use of salicylates, sweats and open bowels. In 1941, Duggan (58) described, under the title of retrobulbar neuritis, what was probably ischemic optic neuropathy, and recommended intravenous injections of sodium nitrite, intramuscular acetylcholine, and inhalations of amyl nitrite.
By the 1930s, most experts had begun to shift their attention from intoxications and infections to primary inflammation of the optic nerve as the potential mechanism of disease. Soon after their introduction in the 1950s, corticosteroids were used to treat optic neuritis. Adrenocorticotrophic hormone (ACTH) became a popular treatment primarily based on the work of Rawson (59). Benefit of ACTH was determined based on near acuity measurements at one month after onset of optic neuritis. However, Giles (60) and Bowden et al. (61) found no significant benefit two years after steroid treatment, but did show that it significantly hastened visual recovery in the first two to three weeks. Bird et al. (62) found that retrobulbar injections of triamcinolone produced faster recovery but no long term differences between treated and untreated patients. In the 1969 edition of their book, Walsh and Hoyt concluded that ACTH was not beneficial (63).
It was not until the last 15 years that the ONTT (64) showed, through a large randomized trial, that oral prednisone treatment, in milligram per kilogram body weight doses, was ineffective and promoted recurrences of the optic neuritis. That trial confirmed that intravenous corticosteroids (methylprednisolone 1 gm/day) for three days, followed by prednisone (mg/kg body weight/day) for eleven days, accelerates visual recovery but does not improve final visual outcome.
I am grateful for the helpful suggestions of Dr. Jeffrey Odell.
Address correspondence to Nicholas J. Volpe, MD, Scheie Eye Institute, 51 North 39th Street, Philadelphia PA 19104; E-mail: firstname.lastname@example.org
Copyright © 2001 Lippincott Williams & Wilkins