Clinical manifestations

Presentation and course

Leber hereditary optic neuropathy is a maternally inherited optic neuropathy typically characterized by rapid visual loss, progressing over days to weeks in 1 eye and days to months later in the other eye. Acute bilateral disease is seen in about one fourth of patients (42). There may be disparity in visual function between the 2 eyes. Although Leber hereditary optic neuropathy is usually painless, some patients may have discomfort reminiscent of optic neuritis (42).

The typical vision loss pattern is decreased visual acuity, with most patients seeing 20/200 or worse (100; 99), a visual field defect involving the blind spot and central fixation (cecocentral scotoma). In the acute stage, there is a characteristic pseudoedema of the optic disc, which does not leak on fluorescein angiography. Fine telangiectatic vessels are often present at the disc margins.

One of the most peculiar features of Leber hereditary optic neuropathy is that the expected relative afferent pupillary defect (Marcus-Gunn pupil) is not as prominent in patients with unilateral disease. This may reflect preservation of W-like retinal ganglion cells, which are thought to mediate the pupillary response, whereas the X-like and Y-like ganglion cells responsible for visual processing are affected by the disease (52; 135). Another possible explanation is relative sparing of the intrinsically photosensitive retinal ganglion cells that contribute significantly to a sustained component of the pupillary light reflex independent of rod and cone pathways. A pupillographic study comparing 10 severely affected Leber patients with 16 healthy age-matched controls demonstrated relative preservation of the pupillary light reflex in the Leber hereditary optic neuropathy patients compared to controls, despite an estimated loss of over 90% of the retinal ganglion cells by optical coherence topography retinal nerve fiber layer thickness determination (93).

As clinical experience with the disease has increased and genetic confirmation has become more common, it has become apparent that many patients may lack the clinical features described above (138). For example, the visual loss may be indolent. Instead of disc swelling, there may simply be optic atrophy, or even cupping (54; 110). The visual fields may be atypical, eg, bitemporal hemianopsia (138).

Prognosis and complications

The prospect for visual improvement in Leber hereditary optic neuropathy is low, with the specific rate highly dependent on which mitochondrial DNA mutation is present (57). The most dismal prognosis is for patients with the 11778 mutation. In a natural history study of 44 patients enrolled at varying time points since symptom onset, 15% of eyes had improvement, 7% had worsening, and the remainder were stable in 1 or more eyes (74). All improvement occurred within 36 months after onset, and all worsening occurred within 12 months of onset. A metaanalysis revealed that ultimate visual acuities of better than 20/200 are rare (98). The best prognosis is for those with the 14484 mutation, in which partial or full recovery may be seen in as many as 71% of patients (118). An intermediate recovery rate of approximately 22% is seen with the 3460 mutation. In all of these mutations, there may be a latency period of months to years until visual improvement (130). Some clinical characteristics associated with better visual recovery include young age of onset and better visual acuity at nadir. Presence of peripapillary telangiectasia and optic disc hyperemia may indicate poor visual prognosis (92). A nationwide cohort study in Denmark comparing 141 affected Leber hereditary optic neuropathy subjects, 297 carriers, and the general population found a rate ratio of 1.95 for mortality and 7.53 for alcohol related disorders in Leber hereditary optic neuropathy subjects versus the general population with no difference between carriers and the general population (134).

As experience with Leber hereditary optic neuropathy has increased, it has become apparent that other neurologic abnormalities may be associated with it as part of Leber hereditary optic neuropathy plus syndromes. Most commonly, patients have a syndrome indistinguishable from multiple sclerosis (43; 32; 107; 55), particularly in female patients (41). Neuropathologic examination of one such patient with 1448 mutation showed extensive frontal lobe demyelination mixed with cavitary necrosis and CD8-positive T-cell predominant inflammatory infiltrates. Several pathophysiologic possibilities are discussed, one being that Leber hereditary optic neuropathy somehow activates an autoimmune process in the central nervous system with some features of multiple sclerosis (72). Microscopic abnormality of normal-appearing brain tissue was demonstrated using MRI measures including magnetization transfer ratio and mean diffusivity histogram analysis in 10 patients with Leber hereditary optic neuropathy, and these abnormalities were found to be more pronounced in 4 patients with Leber hereditary optic neuropathy and multiple sclerosis-like illness (51).

A United Kingdom-wide prospective cohort study of prevalent cases of multiple sclerosis with Leber hereditary optic neuropathy mitochondrial DNA mutations identified 12 new cases from 11 pedigrees, and 44 cases with this combination were identified from the existing literature. Analysis of the clinical features of this hybrid group indicated a clinical profile that is distinct from that of patients with just Leber hereditary optic neuropathy DNA mutations and also distinct from that of patients with just multiple sclerosis. The hybrid group had features more like multiple sclerosis than Leber hereditary optic neuropathy disease, including female preponderance, multiple episodes of visual loss, and a long interval before the fellow eye is affected (1.66 years on average), as well as features that are more typical of Leber hereditary optic neuropathy than multiple sclerosis, including painless vision loss and failure of visual recovery. The authors conclude that the co-occurrence of multiple sclerosis and Leber hereditary optic neuropathy mitochondrial DNA mutations is likely due to chance, but the resulting clinical profile has a phenotype distinct from either group alone, suggesting a mechanistic interaction (112). The possible genetic and pathogenetic linkage between Leber hereditary optic neuropathy and multiple sclerosis was further investigated in a multinational study group called the MAGNIMS network. They utilized a blinded standardized review of brain MRIs from 30 patients with multiple sclerosis, 31 patients with Leber hereditary optic neuropathy, and 11 patients with symptoms of both whom they dubbed as having “LMS.” All of the patients with LMS had white matter lesions by definition, and 25% of patients with Leber hereditary optic neuropathy had similar brain lesions. Many more females with Leber hereditary optic neuropathy had white matter lesions than males (relative risk 8.3), mirroring the gender ratios for patients with multiple sclerosis (89). A study of 31 unrelated Iranian patients with clinically definite multiple sclerosis with optic nerve involvement and 25 patients with clinically definite multiple sclerosis but no optic nerve involvement showed that none of the patients in either group had mitochondrial DNA point mutations at np 11778, 3460, or 14484 (45). A nationwide cohort study in Denmark comparing 141 affected Leber hereditary optic neuropathy, 297 carriers, and the general population found a rate ratio of 12.89 for affected Leber hereditary optic neuropathy versus general population, which was not elevated in carriers (134). There are also reports of Leber hereditary optic neuropathy with positive aquaporin-4-antibody (29) and, more recently, myelin oligodendrocyte glycoprotein antibody (08).

Other specific neurologic associations include cerebellar ataxia (34; 94), tremor (103), spastic dystonia (90), peripheral neuropathy (103), thoracic kyphosis (103), and primary degeneration of spinal cord dorsal columns (56). The association of spastic dystonia and Leber hereditary optic neuropathy has been documented with a novel mitochondrial mutation, 3697G>A/ND1, a mutation that has also been associated with MELAS syndrome (129). The 14459 mutation is characterized by hereditary dystonia (61; 62). Patients with the mitochondrial mutation at np 14484 have associated migraine with or without aura commonly, possibly as a manifestation of abnormal oxidative phosphorylation that has been demonstrated in migraineurs (25). The Danish national cohort study reported increased rate ratios for affected Leber hereditary optic neuropathy subjects compared with the general population for dementia, epilepsy, and neuropathy, but not migraine (134).

A relatively uncommon complication of Leber hereditary optic neuropathy is a cardiac pre-excitation syndrome, either Wolff-Parkinson-White or Lown-Ganong-Levine syndrome. This can occur in up to 9% of Leber hereditary optic neuropathy patients in Finland or Japan (104; 86). Prolongation of the corrected QT interval may also occur (110). A 48-year-old man with Leber hereditary optic neuropathy and Wolff-Parkinson-White syndrome was found to have myocardial thickening and isolated left ventricular abnormal trabeculation on examination using echocardiography and cardiac magnetic resonance imaging. The patient’s brother also had isolated left ventricular abnormal trabeculation. Isolated left ventricular abnormal trabeculation is frequently associated with respiratory chain disorders according to the authors (31). The Danish national cohort study reported increased rate ratios for affected Leber hereditary optic neuropathy subjects compared with the general population for all heart disease, atherosclerosis, and stroke, but not ischemic heart disease, heart failure, or arrythmia (134).

Clinical vignette

A 22-year-old man awoke with hazy vision in both eyes and difficulty seeing in bright illumination. The vision worsened gradually for about 1 month after onset, by which time he could no longer drive a car. The blur was greater in the central field of either eye than in the periphery. He had not experienced any positive visual symptoms.

He also had been having frequent unilateral headaches for the past 8 to 9 months. It was most often on the left side and was usually worse when on the right side. It felt like pressure that starts behind the eye and in the temple and spread to the back of his head and neck. He could feel the muscles of his neck tighten when that area is involved. The headache lasted about 2 hours and occurred 3 to 4 times weekly. Being in the sun or watching TV for a long interval seemed to precipitate the headache. The pain was steady in quality, and there was some globe tenderness with the headache. He had a long history of headaches, but these seemed to him different than past headaches. The differences included greater localization around his orbit and temple and the high frequency. His past headaches were more holocephalic and somewhat less frequent. He never had nausea or vomiting with any of his headaches (past or present). He remembered having a headache when he awoke with the first vision symptoms, but he could not recall which side it involved.

When this started, he saw a local optometrist. He had trouble getting an appointment with an ophthalmologist, so the first ophthalmologic evaluation was a month and a half after symptom onset, at which time the examination was unrevealing. He was referred for retinal evaluation, and once again a normal dilated fundus examination was described as normal. His measured visual acuity was 20/60 for either eye. Fluorescein angiogram was normal. Two months after symptom onset a neuro-ophthalmologist diagnosed optic nerve damage most likely related to alcohol use.

Prior evaluation included brain and orbit MRI with and without gadolinium that was read as normal. Blood work included complete blood count, B12, folate, thiamine, liver function tests, erythrocyte sedimentation rate, and 24-hour urine collection for heavy metals. All were normal. The hematocrit was slightly low, but the hemoglobin level was normal. The erythrocyte indices were normal.

He had a general physical examination by an internist, and this was said to be normal. Past medical history is unremarkable. Past ocular history is negative. Medications include only thiamine 50 mg twice a day and B-complex vitamins. Social history is significant for the increased drinking (alcohol) pattern since last semester at college. The patient reported a pattern of drinking 5 to 7 drinks in the evening 2 to 3 nights per week. He never missed meals and did not experience vomiting or diarrhea. He had smoked cigarettes and cigars “occasionally.” Family history was extensively known, and there was no history of unexplained poor vision.

Examination. Best corrected visual acuity was 20/200 with the right eye and 20/400 with the left. He identified 11 out of 11 color plates with the right eye and 10 out of 11 with the left (Ishihara, +4.00 sphere). Pupils were equal, round, and reactive to light without relative afferent pupillary defect (measured using an infrared video pupillograph and neutral density filters). Versions and ductions were full. Saccades and pursuit were normal into all fields of gaze. There was no significant nystagmus and were no dissociated eye movements. Palpebral fissures were 11 mm and symmetric. Intraocular pressure by applanation was 14 in both eyes. Exophthalmometry was symmetric. Slit lamp examination was unremarkable. Dilated fundus examination showed the optic discs to be a little full in contour, but normal in color and capillary content with somewhat dilated “telangiectatic” vessels on and around the optic discs, particularly the left one. The retinal arterioles and veins were normal, and the maculae were unremarkable. The retinal periphery was normal in either eye. The veins were not tortuous.

Goldmann perimetry. The central and peripheral isopters were normal in either eye. There were dense central scotomas in both eyes; the maximum density in the right eye was I4e and V4e for the left eye. Also, the diameter of the scotoma was larger for the left eye.

Humphrey 24-2. Central threshold examination also demonstrated highly circumscribed central scotomas in both eyes.

Impression. This appears to correspond most closely to a phenotypic Leber hereditary optic neuropathy, possibly triggered by metabolic stresses related to alcohol consumption. The lack of eye pain bilaterality and circumscribed nature of the papillomacular bundle involvement seems unlikely to represent an inflammatory optic neuropathy.

Biological basis

Etiology and pathogenesis

Genetics. Inheritance of Leber hereditary optic neuropathy susceptibility is maternal, consistent with a mitochondrial genome abnormality as the cause (96). Each mitochondrion contains 2 to 10 copies of a closed circular double-stranded DNA coding for 13 of the 67 proteins making up the mitochondrial respiratory chain, as well as the transfer RNA and ribosomal RNA needed for mitochondrial protein synthesis. In most pedigrees, mutations at positions 11778, 3460, or 14484 of the mitochondrial genome are found, and these mutations are not found in pedigrees of unaffected individuals. In general, these mutations involve mitochondrial DNA complex I, or NADH:ubiquinone oxidoreductase (ND) genes (G11778A in ND4, G3460A in ND1, and T14484C in ND6).

In a minority of cases, other mutations have been described including over 7 in the mitochondrial DNA ND6 gene at A14495G and 14568 (22; 30) and at 11253 in the ND4 gene (79). In addition, a mutation at position 14459 is in association with hereditary spastic dystonia with wide variability of clinical expression (61; 62; 126; 36; 132). A mutation was found at G13513 and at 3376 (G> A) in the MTND1 gene with MELAS overlap syndromes (114; 09). These are called primary mutations.

A mutation at position 15257 has been suggested to cosegregate with the disease, but the presence of this mutation in normal subjects has led to controversy about whether this is actually a primary mutation (60; 108; 14; 80). However, the fact that subjects may harbor any given mutation but not develop clinical Leber hereditary optic neuropathy makes it difficult to determine whether there is a causal relationship between the genotype and phenotype for this and similar mutations.

Mutations at other points in the mitochondrial genome may be associated with Leber hereditary optic neuropathy in conjunction with one of the primary mutations. These include 3394, 4160, 4216, 4917, 5244, 7444, 13708, and 15812 (15; 14). It is controversial whether the presence of secondary mutations affects the prognosis of the disease (102). Mitochondrial haplotype grouping reveals that haplogroup J is more often seen in patients with the 14484 and 11778 mutations, suggesting a polygenic effect on disease penetrance (13; 75). In a study of 3613 subjects from 159 Leber hereditary optic neuropathy pedigrees, it was found that vision loss was more prevalent when the primary mutations were found on specific haplotype subgroups—J2 for 11778G to A, J1 for 14484T to C, and haplotype K for 3460G to A (47). In Chinese pedigrees with Leber hereditary optic neuropathy bearing the primary G11778A mutations, secondary mutations at the tRNAMet A4435G and the tRNAThr A15951G loci were found to increase the penetrance of clinical disease (81; 117). Cells from matrilineal family members carrying both the G11778A and the secondary A15951G mutation had about a 35% reduction in the level of tRNAThr relative to those carrying only the G11778A mutation (81). The authors speculated that this reduction of tRNA metabolism could cause impaired mitochondrial translation in subunits including ND4 of NADH dehydrogenase (complex I).

Male predominance is strong in Leber hereditary optic neuropathy, the degree depending on the specific mutation. The genetic basis for this is controversial. It was initially believed that digenic inheritance, with the second gene on the X chromosome, could account for the higher proportion of symptomatic males among obligate carriers with a primary mutation. However, detailed genetic studies argue against an X-linked locus for male susceptibility, and this issue remains a topic of investigation (20; 111). At least 1 pedigree has been described in which 9 patients in 4 generations with the 11778 mutation developed Leber hereditary optic neuropathy, and 8 of the 9 were girls (133).

Although most patients with Leber hereditary optic neuropathy are homoplasmic for a primary mitochondrial DNA mutation, some patients (4% to 14%) may have heteroplasmy or a variable proportion of mutant and wild-type mitochondrial DNA (42; 88). The degree of heteroplasmy probably influences the likelihood of developing symptomatic Leber hereditary optic neuropathy in those at risk (127; 42; 88; 131), but not necessarily the degree of visual loss (127). In an analysis of 17 independent pedigrees harboring the G117678A mutation, it was found that the frequency of blindness in males was related to the mutation load in blood cells and that mothers with 80% or less mutant mtDNA in their blood cells were less likely to have clinically affected sons than mothers with 100% mutant mtDNA in the blood (21). A study of 167 genealogically unrelated Leber hereditary optic neuropathy families revealed that the prevalence of heteroplasmy was 5.6% for the 11778 mutation, 40% for the 3460 mutation, and 36.4% for the 14484 mutation. The authors conclude that this significant variance in prevalence of heteroplasmy between the different primary Leber hereditary optic neuropathy mutations suggests genotypical differences in disease expression (53).

Pathophysiology. The etiology of Leber hereditary optic neuropathy is unclear. Light and electron microscopic studies of Leber hereditary optic neuropathy eyes and optic nerves demonstrate loss of the retinal ganglion cell layer and optic atrophy (65). However, the nature of pathological study prevents determination of the pathophysiology of the disease, for example, whether axonal loss causes (or is caused by) ganglion cell loss or even whether functional mitochondrial abnormalities in the ganglion cell are responsible, as opposed to another cell type (eg, astrocytes or oligodendrocytes).

Mitochondrial dysfunction. Studies of the effects of mitochondrial DNA mutations on mitochondrial function suggest an oxidative phosphorylation defect resulting from the genetic defect. In general, mutations at the 11778 position result in minor changes in oxidative phosphorylation but may affect binding of complex I to ubiquinone (26). Mutations at the 3460 position result in greatly decreased complex I activity (85; 128; 23). There is some evidence that nuclear genome variation alters the expression of mitochondrial complex I expression in persons bearing the 3460 mitochondrial mutation (23). Mutations at the 14484 position result in decreased complex I electron transfer activity and ATP synthesis (109). In vivo imaging using 31P magnetic resonance spectroscopy demonstrates moderately decreased energetics in the skeletal muscle of patients with the 11778 mutation (05) but less so in patients with other mutations (83).

In an attempt to further define the effects of specific mitochondrial DNA mutations on mitochondrial function, several workers have fused rho0206 cells lacking mitochondria with mitochondria from affected individuals, producing cell lines ("cybrids") containing mutated mitochondria. Cybrids with the 11778 mutation have a 40% decrease in NADH dehydrogenase-dependent respiration but not rotenone-sensitive NADH dehydrogenase activity (44), whereas cybrids with the 14459 mutation have a 39% decrease in complex I activity (62). In at least 1 study, however, these metabolic consequences could not be confirmed, and studies using cybrids will probably require further study (18).

In a study using in vivo phosphorus magnetic resonance spectroscopy to measure oxidative phosphorylation in occipital lobe and calf muscle, it was found that in 3 members of a family with the 3460 mutation indices of brain energy metabolism was abnormal in all 3, but muscle oxidative phosphorylation rate was normal in all. This study indicates that the distribution of the biochemical expression of the mutation can be studied in vivo and that it is not equal in all organs of affected persons (82).

Oxidative stress. It has been presumed that abnormal energy production leads to accumulation of reactive oxygen species in retinal ganglion cells that are responsible for the optic neuropathy, resulting in the clinical and pathological findings. Similarities between Leber hereditary optic neuropathy, vitamin B12 deficiency, and nutritional amblyopia have suggested that abnormalities of ATP levels might be causative (119). However, abnormalities of retinal ganglion cell energy production or free radicals have not been demonstrated in an in vitro or in vivo model of Leber hereditary optic neuropathy; thus, it remains the subject of active laboratory investigation. In a study of various nonenzymatic antioxidants and lipid peroxides in the blood of patients with Leber hereditary optic neuropathy carriers with homoplasmic DNA mutation at 11778, it was found that these carriers had lower levels of the alpha-tocopherol/cholesterol + triglyceride ratio than controls without these mutations. It was suggested that alpha-tocopherol may be the primary scavenger molecule against the free radicals induced by complex I deficiency. The authors also suggested that the reduced alpha-tocopherol levels resulted from consumption of the antioxidant by the affected tissues (67). Wang and coworkers assessed free radicals in venous blood from 14 patients, 20 asymptomatic relatives, and 30 control subjects using luminal luminescence after addition of phytohemagglutinin (137). They found that free radicals increased in the blood of Leber hereditary optic neuropathy patients and their asymptomatic relatives to a much greater extent than in normal controls, indicating reduced antioxidant capacity in members of the affected families. An optic neuropathy that is pathologically similar to that of Leber hereditary optic neuropathy has been produced in a DBA/1J mouse model of oxidative injury by increasing mitochondrial levels of reactive oxygen species (115).

Howell has suggested that respiratory chain dysfunction might lead to axoplasmic flow stasis at the level of the cribriform plate, with secondary damage from swelling of axons in that confined space, which he calls a “chokepoint” (46).

Serum neuron-specific enolase is a biomarker for neuronal stress. Serum neuron-specific enolase levels from 74 members of a pedigree with Leber hereditary optic neuropathy and homoplasmic 11778/ND4 mitochondrial DNA mutation demonstrated significantly higher levels in 46 asymptomatic carriers (27.17±39.82 µg/L) than in 14 symptomatic “affected” members (5.66±4.19 µg/L; p=0.050) and 14 “off-pedigree” controls (6.20±2.35 µg/L; p=0.047) (139). Among the carriers, levels were much higher in males (40.65±51.21 µg/L) than in female carriers (15.85±22.27 µg/L; p=0.034).

Penetrance determinants. Beyond contributions of secondary mutations and haplotypes discussed above, the fact that genetically identical monozygotic twins may be discordant for long periods of time for Leber hereditary optic neuropathy implies that epigenetic factors affect the susceptibility to the disease (59; 73). Caporali and colleagues provide an excellent review on this topic (16).

Kirkman and colleagues have shown that oxidative stress is strongly associated with the onset of symptomatic vision loss in a study of 196 affected and 206 unaffected members of 125 pedigrees that harbor 1 of the 3 primary mitochondrial DNA mutations (66). They found a strong and consistent association between vision loss and smoking, with penetrance of 93% among males who smoked. There was a lesser trend toward increased vision loss with alcohol ingestion but only with very heavy intake.

Another factor affecting susceptibility to clinical manifestations may be the possibility of adaptive change in the mtDNA in the presence of point mutations. One study investigated the quantitative ratio of mtDNA to nuclear DNA in peripheral circulating leukocytes from 13 asymptomatic carriers and 18 family noncarriers related to 11 patients with Leber hereditary optic neuropathy due to the 14486 mutation. Significant increase in the mtDNA relative to nuclear DNA was found only in asymptomatic carriers, indicating that those who had not increased the mtDNA had become symptomatic (105). Another study found that increase in cellular mtDNA content may protect against symptomatic conversions in individuals with heteroplasmic 11778 and 3460 mutations (07).

Asymptomatic manifestations. In a psychophysical study of 18 asymptomatic carriers of the 11778 mutation versus 18 control subjects, Gualtieri and colleagues demonstrated abnormally high contrast discrimination thresholds among the carriers as compared with normal control subjects (37). This suggests that even in asymptomatic carriers there are subtle abnormalities of visual processing that may be used to identify them. Decline in pattern electroretinogram measurements have been reported in asymptomatic carriers over time (38).

Microvasculopathy. The advent of optical coherence tomography angiography has improved visualization and quantitative measurement of optic nerve and retinal microvasculature in Leber hereditary optic neuropathy. Peripapillary microvascular changes correlate with ganglion cell–inner plexiform layer loss and precede retinal nerve fiber layer thinning (03). Microvascular attenuation was also reported in the macular region corresponding with papillomacular bundle (11). These microvascular changes may play an integral part of Leber hereditary optic neuropathy pathophysiology.

Epidemiology

Leber hereditary optic neuropathy is an uncommon cause of optic neuropathy. Its prevalence is highly dependent on the population studied, as well as which mutation is being analyzed. In Australia 2% of those disabled by blindness have Leber hereditary optic neuropathy as a cause (84).

The 11778 is the most prevalent mutation, with the proportion varying depending on population. In Japan, approximately 80% to 90% of patients with Leber hereditary optic neuropathy have the 11778 mutation, whereas 40% to 85% of non-Japanese patients have that mutation (95; 52).

The age of onset is usually in the second or third decade but has been reported as early as 4 years of age (91) and as late as 73 years of age (01), with many cases over 60 years of age at time of symptom onset described in the literature (113; 27). The age of onset does not differ significantly between the different mutations (42).

Male predominance is strong in patients with Leber hereditary optic neuropathy (99), the degree depending on the specific mutation. In a European population, there were male-to-female ratios of 3.7:1 with the 11778 mutation, 4.3:1 with the 3460 mutation, and 7.7:1 with the 14484 mutation (42). In a study comparing 16 women and 66 men with Leber hereditary optic neuropathy, it was found that women were older at presentation (average 31.3 vs. 24.3 years), had more severe vision loss, less tendency to recover vision, and a much higher rate of having an affected mother than did affected men (78).

The likelihood that those genetically at risk will develop visual loss depends on gender. The likelihood of symptomatic disease affecting matrilineal first-degree relatives of affected individuals is approximately 20% to 46% for male relatives and 4% to 10% for female relatives (84; 42).

Prevention

The most important risk factor for development of Leber hereditary optic neuropathy is the presence of one of the primary mutations (11778, 3460, and 14484). As the mitochondrial DNA are maternally inherited, it can be expected that the children of an affected female, but not of an affected male, will harbor the mutation. Similarly, a cousin linked through a female lineage to an affected subject may be at risk, as would any similar relative. DNA testing can confirm whether or not an individual relative has the relevant mutation.

No prophylaxis has been convincingly shown to be of value in preventing development of Leber hereditary optic neuropathy in those genetically at risk. Because of the presumption that mutations in genes coding for respiratory chain subunits result in functional abnormalities of oxidative phosphorylation, some clinicians suggest that their patients take vitamin C, vitamin E, coenzyme Q10, or other antioxidants. For similar reasons, many patients are counseled to avoid use of tobacco or alcohol. Some suggest avoiding foods containing naturally occurring cyanide, which interferes with mitochondrial respiration. A systematic epidemiologic and neuro-ophthalmologic study of a large Brazilian pedigree with 11778 haplogroup J mutation has demonstrated a strong influence of environmental risk factors in the development of phenotypic disease, particularly smoking (123).

Differential diagnosis

Confusing conditions

Leber hereditary optic neuropathy is often confused with other optic neuropathies and may even be misdiagnosed as functional visual loss. The presence of a family history, especially in the maternal lineage, is helpful in making the diagnosis but is not always known or present. Instead, certain clinical observations may help in distinguishing Leber hereditary optic neuropathy from other optic neuropathies.

Other hereditary optic neuropathies may be distinguished based on a combination of clinical features and inheritance pattern. Kjer dominant optic atrophy has an autosomal dominant mode of inheritance mapped to chromosome 3q. Typically with onset from 4 to 8 years of age, it is slowly progressive and rarely results in visual acuity worse than 20/200. Although temporal disc cupping and a tritan (blue-yellow) color defect are typical, these findings may also be seen in Leber hereditary optic neuropathy (54). Recessive optic atrophy is a severe, usually congenital hereditary optic neuropathy associated with visual acuity worse than 20/200, nystagmus, and achromatopsia. The age of onset usually distinguishes it from Leber hereditary optic neuropathy. Other hereditary optic neuropathies may be associated with neurodegenerative disorders, such as Charcot-Marie-Tooth disease and Friedreich ataxia.

It is often difficult to distinguish Leber hereditary optic neuropathy from other optic neuropathies when only one eye is involved. In these cases, the visual loss and disc elevation of Leber hereditary optic neuropathy may be mistaken for optic neuritis (papillitis), anterior ischemic optic neuropathy, or anterior compressive, infiltrative, or infective optic neuropathies. Optic neuritis is typically associated with ocular discomfort aggravated by eye movement, which is only occasionally seen in Leber hereditary optic neuropathy (42). Recovery of vision from optic neuritis usually begins after 1 or 2 weeks, and substantial improvement is often seen by 6 weeks, unlike the usually persistent visual loss of Leber hereditary optic neuropathy. High signal abnormalities on T2-weighted MRI images of the deep white matter are often seen in patients with optic neuritis, but a multiple sclerosis-like syndrome with corresponding neuroimaging findings can also be seen in Leber hereditary optic neuropathy (43; 41; 32; 107; 55).

Nonarteritic anterior ischemic optic neuropathy produces rapid painless visual loss, but is uncommon in patients less than 50 years old and is frequently associated with altitudinal visual field defects corresponding to segmental disc edema. Arteritic anterior ischemic optic neuropathy is seen in elderly patients with symptoms, signs, and laboratory evidence of giant cell arteritis. There is usually pallid disc edema and severe visual loss.

Orbital lesions that compress the optic nerve (eg, extraocular muscle enlargement from Graves disease or orbital tumors), infiltrative disease of the optic nerve (eg, lymphoma, metastatic carcinoma, sarcoid, etc.), or infections of the optic nerve (eg, Cryptococcus or cytomegalovirus) may cause disc edema and visual loss. The time course of these processes is usually less rapid than that of Leber hereditary optic neuropathy, but in some cases, they may overlap. Associated clinical findings, such as proptosis, a history of immunosuppression, or the presence of a known primary cancer, suggest one of these etiologies. However, in some cases the history and examination do not aid in distinguishing these optic neuropathies, and neuroimaging, laboratory studies, and sampling of cerebrospinal fluid are required.

The differential diagnosis of bilateral Leber hereditary optic neuropathy is different. When disc elevation is present without severe visual loss, the possibility of papilledema (ie, disc edema secondary to increased intracranial pressure) should be considered. In papilledema, the visual acuity is initially normal, and the visual field either is normal or demonstrates enlargement of the blind spot. Associated symptoms (headache, vomiting, tinnitus, transient obscurations of vision lasting a few seconds) and signs (unilateral or bilateral sixth nerve palsy, absence of spontaneous venous pulsation at the disc) may suggest the diagnosis. Other causes of bilateral disc edema and visual loss include bilateral presentation of one of the optic neuropathies mentioned above and acute toxic optic neuropathies.

Bilateral Leber hereditary optic neuropathy without disc elevation should be differentiated from nutritional and toxic optic neuropathy, bilateral presentation of a retrobulbar optic neuropathy, low-tension glaucoma (76; 87), and occult retinal dystrophy, particularly a cone dystrophy. All of these may demonstrate decreased visual acuity and central visual field loss. Optic disc pallor may only develop late in the course of the optic neuropathies and may only be present on the temporal disc, where it may be hard to distinguish from normal variation in color. Retrospective surveillance of patients with the diagnosis of tobacco-alcohol amblyopia may yield Leber hereditary optic neuropathy mutations (24). Cone dystrophies may only be detected by electroretinography.

Finally, functional visual loss is a diagnosis of exclusion in a patient with vision loss, no relative afferent pupillary defect, and relatively normal discs. Patients with relatively symmetric vision loss from bilateral optic neuropathy will not have a relative afferent pupillary defect because this test compares conduction in the 2 optic nerves. Because the optic disc pallor of Leber hereditary optic neuropathy may occur late or be difficult to detect, patients should not be diagnosed as having factitious visual loss without considering other possibilities. In some patients, the presence of a constricted "tubular" field (best demonstrated at the tangent screen) instead of the cecocentral scotoma of Leber hereditary optic neuropathy may help in making this diagnosis.

Diagnostic workup

In a symptomatic patient presenting with optic neuropathy, the most reliable way to determine the likelihood of Leber hereditary optic neuropathy is to analyze the patient's mitochondrial DNA for the presence of one of the primary mutations. A peripheral blood sample may be sent to a commercial or research laboratory. Testing usually involves polymerase chain reaction amplification of the appropriate regions of the mitochondrial genome and determination of the presence of a mutation by changes in restriction digest pattern or DNA sequencing, or by allele-specific amplification (106). Results may take days or weeks to return.

To make a clinical diagnosis, visual acuity and visual field testing can identify if the pattern of vision loss is typical of Leber hereditary optic neuropathy. In the acute setting, ophthalmoscopic examination can identify the typical pseudoedema and peripapillary telangectasias. Fluorescein angiography may help determine whether disc elevation represents true edema or the pseudoedema of Leber hereditary optic neuropathy. In the former, but not in the latter, there is leakage of fluorescein from the disc during the late phase of the angiogram. In the chronic stage, ophthalmoscopic examination and ophthalmic imaging will show optic atrophy.

As the differential diagnosis of Leber hereditary optic neuropathy is large, other diagnostic procedures are frequently performed prior to DNA testing. Magnetic resonance imaging with intravenous contrast and fat suppression may demonstrate abnormal enhancement along the optic nerve, for example, as seen in optic neuritis or infiltrative or infectious optic neuropathies but also occasionally seen in Leber hereditary optic neuropathy (64; 12). Additionally, T2 hyperintensity of the posterior part of the affected optic nerve often associated with enlargement of the optic chiasm may also occur in Leber hereditary optic neuropathy (10). Lumbar puncture may detect evidence of central nervous system inflammation, infection, or neoplasm. Serological and radiological studies help in the diagnosis of disorders such as syphilis, vitamin B12 deficiency, and sarcoid. Electroretinography may detect changes consistent with cone dystrophies, although in some cases focal electroretinography may be necessary. Obtaining pattern-shift visual evoked responses (which are abnormal or unrecordable) is usually unnecessary, except in cases of suspected nonorganic visual loss.

There is great interest in identifying presymptomatic changes in genetically affected individuals who are not yet symptomatic as this would offer an opportunity to therapy. Changes in the thickness of the retinal nerve fiber layer can now be measured noninvasively using optical coherence tomography and this has shown characteristic changes in the pre-symptomatic stages as well as after the onset of vision loss. In the months prior to symptomatic visual loss, there is characteristic thickening of the retinal nerve fiber layer in the upper and lower arcuate regions and nasal to the disc with relatively little thickening temporal to the disc in the papillo-macular bundle fibers (125). This correlates with the ophthalmoscopic appearance of pseudo-disc edema in this condition. It is postulated that the thickening relates to dysfunction in energy dependent axoplasmic transport in these ganglion cell nerve fibers. In the acute stage of vision loss, there is rapid thinning in the papillomacular bundle, commensurate with the selective development of central scotoma on visual field examination and with the selective susceptibility of the small diameter axons of the papillomacular bundle (06). Another study examined 6 eyes from 4 patients with Leber hereditary optic neuropathy using spectral domain optical coherence tomography to measure thickness changes in the retinal ganglion cell inner plexiform layer and the macular retinal nerve fiber layer during the presymptomatic stage and at months 1, 3, 6, and 12 after visual loss. Thinning of both the inner plexiform layer and the macular retinal nerve fiber layer was detected in the presymptomatic stage, which may have important implications for therapeutic or preventive interventions even prior to vision loss in 1 eye (04). In a case series of 6 carriers with the 11778G>A mutation followed through conversion from asymptomatic to affected, there was decline in visual acuity, visual field mean deviation, and thickening of the retinal nerve fiber layer on optic coherence tomography in the months prior to symptoms being reported by the subjects (49).

Management

Options for effective treatment of Leber hereditary optic neuropathy have been a long time coming. The successful launch of the antioxidant idebenone followed by its introduction into clinical practice across Europe was an important step forward. Several pharmaceutical agents, acting on different molecular pathways, are currently under development (02).

Karanjia and colleagues offers an excellent review of current therapeutic studies (63). Nonetheless, on the basis of clinical experience suggesting that Leber hereditary optic neuropathy may occur in the setting of tobacco or ethanol abuse, many physicians advise their patients to stop smoking and drinking alcohol. Similarly, because the mitochondrial DNA mutations in this disorder affect subunits of the electron transport chain, some clinicians offer their patients the option of taking vitamin C, vitamin E, and coenzyme Q10. Good glucose control may be of help in patients with diabetes mellitus (28).

Idebenone has been studied for treatment of Leber hereditary optic neuropathy. Idebenone is a synthetic benzoquinone that crosses the blood-brain barrier and mitochondrial membranes. The proposed benefit stems from idebenone’s ability to pass electrons to complex III, bypassing defective complex I in the mitochondrial electron transport chain. Sabet-Peyman and coworkers reported a particularly striking example of visual recovery in a 31-year-old woman with the 11778 mutation whose visual acuity had dropped to 20/200 both eyes when treatment was initiated 2 months after initial vision loss (122). She was given oral idebenone 900 mg/day for 9 months after a 3-day course of both intravenous methylprednisolone at a dose of 1 gram/day and oral coenzyme Q10 at a dose of 200 mg/day. By the ninth month of treatment, the visual acuity had improved to 20/25 for both eyes with marked improvement in the central visual field. The retinal nerve fiber layer was thicker than normal as usual during the acute stage, and at 9 months, thickness had returned to normal, and there was no pathologic thinning of the nerve fiber layer (122). In a retrospective study of 103 patients with 1 of the 3 main mitochondrial gene mutations in Leber hereditary optic neuropathy, 44 were treated with idebenone in doses ranging from 270 to 675 mg/day, but treatment was initiated within the first year of visual loss onset in all (17). Benefit of treatment reached statistical significance only in the group with the 11778 mutation, probably because of the higher rate of spontaneous visual improvement in those with the 14484 mutation.

A prospective, randomized, placebo-controlled trial involved treatment of 84 patients with Leber hereditary optic neuropathy and 1 of the 3 primary mitochondrial gene mutations (68). Fifty-five patients were treated with idebenone 900 mg/day, and 30 patients received placebo. Treatment was carried on for 6 months and was initiated at various times in the course of the disease, sometimes years after first vision loss. This study failed to reach statistical significance for its primary endpoint—best recovery of visual acuity—but did reach significance in some secondary endpoints and subgroup analyses. Further study of color contrast sensitivity in 39 Leber hereditary optic neuropathy patients from a prospective treatment study, 28 taking idebenone and 11 taking placebo, demonstrated a prominent tritan defect and a somewhat less prevalent protan defect (120). Treatment with idebenone improved both color defects compared with placebo, but the tritan defect responded more robustly than the protan defect, and the preservation of color vision was most prominent in patients who had had vision loss mainly in one eye, indicating an important protective effect of idebenone against development of color vision deficiency.

An open-label, multicenter, retrospective, noncontrolled analysis of long-term visual acuity and safety in 111 Leber hereditary optic neuropathy patients treated with idebenone (900 mg/day) demonstrated the benefit of idebenone treatment in recovering vision loss and maintaining residual vision (19). The authors suggested that idebenone treatment should be initiated early and be maintained more than 24 months to maximize efficacy. The safety profile was consistent with known data.

RESCUE is a multicenter, randomized, double-masked, sham-controlled, phase 3 clinical trial that evaluated the efficacy of a single intravitreal injection of rAAV2/2-ND4 in subjects with visual loss from Leber hereditary optic neuropathy (101). At 96 weeks after unilateral injection, 1778G>A mutation carriers treated within 6 months after vision loss achieved comparable visual outcomes in the injected and uninjected eyes.

REVERSE is a randomized, double-masked, sham-controlled, multicenter, phase 3 clinical trial that evaluated the efficacy of a single intravitreal injection of rAAV2/2-ND4 in subjects with visual loss due to Leber hereditary optic neuropathy (140). A total of 37 subjects carrying the m.11778G>A (MT-ND4) mutation and with duration of vision loss between 6 to 12 months were treated. Each subject's right eye was randomly assigned in a 1:1 ratio to treatment with rAAV2/2-ND4 (GS010) or sham injection. The left eye received the treatment not allocated to the right eye. Unexpectedly, sustained visual improvement was observed in both eyes over the 96-week follow-up period. At week 96, 25 subjects (68%) had a clinically relevant recovery in BCVA from baseline in at least one eye, and 29 subjects (78%) had an improvement in vision in both eyes. Evidence of transfer of viral vector DNA from the injected eye to the anterior segment, retina, and optic nerve of the contralateral noninjected eye supports a plausible mechanistic explanation for the unexpected bilateral improvement in visual function after unilateral injection (140).

EPI-743 is a third-generation ubiquinone that has been shown to exhibit approximately 1000 times greater in vitro activity as idebenone as an antioxidant. In an initial open-label study of 5 patients with vision loss from Leber hereditary optic neuropathy, treatment with oral EPI-743 was initiated at various intervals after onset of vision loss and was continued for at least a year in all (124). Significant objective and subjective stabilization and improvement of various visual parameters occurred in 4 of the 5 patients. These encouraging results have prompted the authors to develop an international multicenter prospective controlled trial using EPI-743 to treat patients with Leber hereditary optic neuropathy. Other compounds are currently being studied.

Topical application of brimonidine purite 4 times a day to the unaffected eye in 9 patients with visual loss in 1 eye and a primary mitochondrial gene mutation failed to prevent eventual visual loss in the second eye (97).

A radically different approach to management via gene therapy may be on the horizon for patients who have already lost vision in 1 or both eyes as well as for those who have 1 of the primary mitochondrial gene mutations but have not lost the vision of either eye. Koilkonda and Guy provide a comprehensive and thoughtful overview of this prospect in their important review article (69). Although exogenous genes have been injected into the nuclear genome to reverse genetic mutation effects, this has not been possible to do with the mitochondrial genome. To bypass this difficulty, these workers used a process called allotopic expression, in which a nuclear-encoded version of the ND-4 subunit of Complex I gene that is normally encoded by mtDNA is injected into the cell nucleus. The base pair sequence that encodes ND-4 in the nuclear genome varies somewhat from that of the mtDNA encoded ND-4. The protein product of this nuclear ND-4 gene is expressed in the cytoplasm and imported into the mitochondria by appending a mitochondrial transport peptide on the N-terminal and a FLAG tag on the C-terminal by which to identify the gene product.

To study the effectiveness of allotopic expression, Guy and colleagues created transmitochondrial hybrid cell lines (cybrids) by fusing enucleated patient cells homoplasmic for wild type (11778G) or mutant (G11778A) mitochondrial DNA with neutral nucleated host cells that have permanently lost all mitochondrial DNA after exposure to ethidium bromide (40). They then showed that in transmitochondrial cybrids with G11778A mutated mitochondrial DNA, ATP synthesis dependent on complex I substrates was substantially reduced and that this deficiency in oxidative phosphorylation can be reversed using allotopic expression of the ND-4 gene and transport of the gene product into the mitochondria that are still homoplasmic for G11778A mutant mtDNA (116).

Taking these techniques another step closer to human intervention, Guy and colleagues injected a nuclear encoded ND-4 subunit fused with a mitochondrial targeting sequence on the N-terminal and a FLAG epitope on the C-terminal for subsequent identification into the right eyes of mice using an adeno-associated viral plasmid vector (39). As a control, the left eyes of the mice were injected with adeno-associated viral plasmid vector tagged with green fluorescent protein. They were then able to show that the human ND-4 was effectively integrated into the murine Complex I within mitochondria of the retinal ganglion cells and optic nerve axons and that expression of human ND-4 did not induce loss of retinal ganglion cells, reduction of ATP synthesis, or reduction of pattern electroretinogram amplitude. Thus, they did not demonstrate any impediment to further consideration of human treatment using this technique (70). This safety study was extended to additional animals, including nonhuman primates, and increased the ratio of mutant to wild-type ND-4 gene administered by intravitreal injection without any reduction in the prevention of optic neuropathy and visual loss (71).

There are 3 groups studying this approach in humans based in the United States, France, and China. A prospective open-label trial in which the study drug, a self-complementary adeno-associated virus vector expressing a normal ND4 complementary DNA, was intravitreally injected unilaterally into the eyes of 5 blind patients with G11778A Leber hereditary optic neuropathy (33). Four of these had visual loss for more than 12 months and the fifth patient for less than 12 months. Study subjects were followed for 90 to 180 days with ocular and systemic safety assessments and visual examinations. Visual acuity on the Early Treatment Diabetic Retinopathy Study (ETDRS) eye chart remained unchanged in the first 3 participants. At 90 days follow up visual acuity, 2 subjects had improved from hand motion vision by 7 letters, and in 1 patient, it had improved by 15 letters, which is the equivalent of 3 lines. No participant lost vision, and no serious adverse events were documented. A Chinese study studied 6 patients over 9 months with some suggestion of improvement in visual function and no safety concerns. Visual function index and baseline best corrected visual acuity were found to be independent prognostic factors in a Leber hereditary optic neuropathy 11778 patient receiving ND4 gene therapy. Logistic regression showed that better baseline visual function index and best corrected visual acuity correlated with better visual prognosis (141). Interestingly, bilateral visual acuity improvement has been observed in 29 subjects (78%) out of 37 unilaterally treated subjects from the REVERSE study. Mean improvement in best-corrected visual acuity at week 96 was also similar between recombinant AAV-treated eyes and sham-treated eyes (140).

Outcomes

A meta-analysis of 12 retrospective and 3 prospective studies reviewed visual function of 695 Leber hereditary optic neuropathy patients with the m.11778G>A mutation (101). One hundred (14.4%) patients reported vision recovery, although idebenone use in some studies could not be excluded. The m.11778G>A mutation of severe visual loss with rare or poor recovery from nadir still holds true for most affected individuals. In patients aged 15 years or older, meaningful visual recovery occurred in 23 of 204 (11.3%) and ultimate visual acuities of better than 20/200 were rare. Patients younger than 12 years of age portended a better visual prognosis and a different natural history of visual loss progression and recovery.