Progressive Visual Recovery in a Patient with Leber Hereditary Optic Neuropathy Harboring a Rare Heteroplasmic (MT-ND5):m.13042G>A, p(Ala236Thr) Variant with Low Mutant Load: A Case Report

Abstract

Introduction

The majority of Caucasian patients with Leber hereditary optic neuropathy (LHON) harbor one of three primary pathogenic mitochondrial DNA (mtDNA) variants, which are usually present in homoplasmy in leukocytes. Most of the remaining cases have been linked to rare heteroplasmic pathogenic variants, which typically require a mutational load of more than 60% to result in phenotypic expression.

Case Presentation

An 18-year-old Caucasian man presented with sudden visual loss in the left eye. Eye examination revealed hyperemic optic discs with retinal vascular tortuosity and subtle peripapillary telangiectasia, reminiscent of LHON, but initial mtDNA analysis was negative. Best corrected visual acuity (BCVA) continued to decrease despite systemic corticotherapy and subsequent treatment with plasma exchange. The patient then experienced visual loss in the fellow eye, with BCVA in both eyes deteriorating to “counting fingers”. Subsequent screening of complete mtDNA by massive parallel sequencing of leukocyte DNA identified a variant NC_012920.1(MT-ND5):m.13042G>A, p.(Ala236Thr), in heteroplasmy, with a variant load of around 23.8%. Progressive visual recovery was observed in both eyes, resulting in BCVA of 20/29 in both eyes after 4 years of follow-up.

Conclusion

In patients with a strong clinical suspicion of LHON, complete mitochondrial genome sequencing should be considered when initial testing, typically limited to the three primary mutations, is negative. Furthermore, the diagnosis of LHON should not be dismissed if “low” blood mutant loads are found, as important discrepancies of heteroplasmy levels between different tissues have been reported for variants located in the mitochondrial ND5 gene.

Introduction

Leber hereditary optic neuropathy (LHON) is an inherited disease caused by maternally transmitted pathogenic mitochondrial DNA (mtDNA) variants and is characterized by painless, (sub)acute, bilateral (simultaneous or sequential) visual loss due to the preferential loss of retinal ganglion cells within the inner retina, resulting in optic nerve degeneration [1].

The vast majority (90–95%) of patients of Caucasian origin with LHON harbors one of three primary pathogenic mtDNA point variants: m.11778G>A (69%), m.14484T>C (14%), and m.3460G>A (13%). These variants all involve genes encoding the NADH-ubiquinone oxidoreductase (ND) subunits of complex I of the mitochondrial respiratory chain (affecting the ND4, ND6, and ND1 subunits, respectively), leading to selective degeneration of retinal ganglion cells [2–5]. These three primary pathogenic variants are usually homoplasmic (every mtDNA molecule harbors the mutant molecule), but 10–15% of individuals who harbor a primary LHON variant are thought to be heteroplasmic (mutation present in only a fraction of the mtDNA molecules) [1, 2, 4–7].

The remaining cases of LHON have been linked to several other (relatively rare) mutations, typically found in the mitochondrial encoded ND1, ND4, and ND6 genes, which are considered mutational “hotspots” [7]. These mutations are usually heteroplasmic, and most often only have been reported in single patients or individual families [5–7]. These variants are considered pathogenic if impaired protein and mitochondrial function can be confirmed by proteomic and mitochondrial function analysis [5, 7].

We report a case of LHON-like optic neuropathy presenting with acute bilateral sequential visual loss, followed by good, progressive visual recovery. In this patient, we were able to identify a heteroplasmic mtDNA variant NC_012920.1(MT-ND5):m.13042G>A, p(Ala236Thr).

Case Presentation

An 18-year-old male patient of Caucasian origin presented with a 1-week history of sudden visual loss in the left eye. Personal medical history and family history regarding ocular disease were unremarkable. At the time of presentation, he did not take any medications and had not undergone any recent vaccinations. He was only an occasional alcohol user but had been smoking a pack of cigarettes every 2 or 3 days for the past 2 years. Best corrected visual acuity (BCVA) at presentation was 20/20 in the right eye and 20/40 in the left eye, the latter associated with an incipient paracentral scotoma on Humphrey 24-2 automated visual field testing and altered color vision (using Ishihara pseudo-isochromatic plates initially and later confirmed with both Farnsworth-Munsell D15 and Lanthony Desaturated D15 Color Vision Tests) (Fig. 1a, b). Eye movements were full but somewhat painful in the left eye. Pupil size was symmetric with good response to light bilaterally. A subtle relative afferent pupillary defect could, however, be found in the left eye. Slit lamp examination showed no signs of intra-ocular inflammation. Eye fundus examination showed subtle hyperemia of both optic discs as well as subtle tortuosity of the blood vessels (Fig. 2a, b). Spectral-domain optical coherence tomography revealed both slightly increased thickness of the peripapillary retinal nerve fiber layer (RNFL) and slightly decreased thickness of the macular ganglion cell layer in the left eye compared to the right eye (Fig. 3). Fluorescein angiography showed no optic disc hyperfluorescence nor any signs of vascular or chorioretinal inflammation. The presence of small telangiectasia at the border of both optic discs could, however, be suspected (Fig. 2c, d).

Fig. 1.

Humphrey 24-2 automated visual field test demonstrating evolution of visual field defects at initial presentation (a, b), at onset of visual loss in the right eye after 3 months of follow-up (c, d), at nadir of visual field loss in both eyes 2 months later (e, f), and finally at 3 years of follow-up (g, h) (with no more improvement between 3 and 4 years of follow-up). L corresponds to the left eye and R corresponds to the right eye.

Fig. 2.

Eye fundus photography at initial presentation (a, b) demonstrating subtle hyperemia of the optic disc as well as subtle tortuosity of the retinal blood vessels in both eyes, with mid-phase fluorescein angiography (c, d) confirming the presence of small telangiectasia (white arrows) at the border of both optic disc, in the absence of optic disc hyperfluorescence and/or signs of vascular or chorioretinal inflammation (with a, c corresponding to the right eye and b, d corresponding to the left eye).

Fig. 3.

Heidelberg spectral-domain optical coherence tomography at initial presentation demonstrating slightly increased thickness of the peripapillary retinal nerve fiber layer (RNFL) in the left eye (b) compared to the right eye (a), as well as slightly decreased thickness of the macular ganglion cell layer in the left eye (d) compared to the right eye (c).

Comprehensive workup for acute optic neuropathy was initiated, and corticosteroid therapy was introduced for possible optic neuritis. Initial blood tests showed slightly elevated levels of serum angiotensin converting enzyme (63 U/L ‑ normal values 8–55 U/L) as well as for serum lysozyme (20 μg/mL – normal values 9.56–17.14 μg/mL). Calcium level was normal. Blood sedimentation rate was 10 mm/h (normal values <15 mm/h) and C-reactive protein (CRP) 1.2 mg/L (normal values <5.0 mg/L). Antinuclear antibodies, antineutrophil cytoplasmic antibodies, anti-aquaporin 4 antibodies, and anti-myelin oligodendrocyte glycoprotein antibodies were all negative. QuantiFERON assay for tuberculosis and serology for Treponema and Borrelia burgdorferi were also negative. A blood sample was also obtained to look for a pathogenic variant in LHON-implicated genes. Cerebral magnetic resonance imaging showed a slightly hyperintense appearance of the proximal portion of the left optic nerve on T2 fluid-attenuated inversion recovery sequences but without enhancement after contrast injection. Because initial presentation was at the height of the (local) fifth wave of the COVID-19 epidemic and the patient (who had not been vaccinated against COVID-19) had some minor complaints of a cough and a sore throat, COVID-19 screening was also performed, resulting in a positive PCR test for Omicron variant BA.1.

Unfortunately, despite systemic treatment with methylprednisolone 1 g/day intravenous pulses for 3 consecutive days, BCVA in the left eye deteriorated to 20/100 and later to 20/200 despite another three pulses of intravenous corticosteroid therapy. Three weeks after initial presentation, the patient presented at the emergency department with painful swelling of the left parotid gland. Blood test showed elevated levels of CRP (43.0 mg/L) and white blood cells 12.61 × 10³/µL (normal values 3.50–11.00 × 10³/µL) with elevated neutrophils 9.71 × 10³/µL (normal values 1.50–6.70 × 10³/µL) and monocytes 1.49 ×10³/µL (normal values 0.20–1.00 ×10³/µL). Serology for mumps, Epstein-Barr virus, cytomegalovirus, and varicella zoster were all negative. Symptoms resolved within a few days with combined amoxicillin and clavulanic acid antibiotic therapy, with CRP levels rapidly dropping within normal values. Additional chest radiography and positron emission tomography scan showed no evidence for sarcoidosis.

Meanwhile, mtDNA analysis focusing on LHON genotypes (as tested on peripheral blood leukocytes) had shown no evidence for the presence of any of the three primary pathogenic variants. Because of persistent clinical suspicion, sequencing of the complete mitochondrial genome was undertaken in search of any other pathogenic variant implicated in LHON. Since an underlying inflammatory disease could still not definitively be ruled out either, the possibility of a second-line anti-inflammatory treatment was also discussed with the patient. He agreed to undergo plasma exchange (treatment sessions every other day for 6 sessions in total); but unfortunately, follow-up 1 week after the last session once again showed visual deterioration, with BCVA in the left eye now evaluated at 20/320.

Roughly 3 months after initial presentation, the patient also started to experience visual loss in the fellow eye (20/100), with BCVA in both eyes eventually deteriorating to “counting fingers”. Repeated automated visual field testing showed progressive central and paracentral temporal scotoma in both eyes (Fig. 1). The patient was supplied with all our available samples of idebenone (corresponding to a treatment duration of approximately 1 month at 900 mg per day). Sequencing of mtDNA of leukocytes finally identified a heteroplasmic variant m.13042G>A, p(Ala236Thr) in the MT-ND5 gene, with a variant load of around 23.8%. Subsequent genetic testing on peripheral blood leukocytes of the patient’s mother showed no evidence of the same mutation (with a detection limit around 1% for this specific position). Additional testing on the patient’s mother’s urinary epithelium (to account for possible discrepancies of heteroplasmy levels between different tissues) was proposed afterward, but she failed to show up for her appointment and did not contact us again. The proband himself also did not undergo additional genetic testing on other tissues (such as urinary epithelium or skeletal muscle).

However, despite the limited duration of idebenone intake (no more access to samples and only local governmental reimbursement for the three primary pathogenic variants), BCVA in the left eye of our patient started improving to 20/100 roughly 9 months after disease onset, 6 months later followed by the right eye, with BCVA 20/125 in the right eye and 20/50 in the left eye at 15 months of follow-up. Two years after initial presentation, BCVA was evaluated at 20/100 in the right eye and 20/33 in the left eye, a year later, BCVA had improved to 20/33 in the right eye and to 20/29 in the left eye, and 1 more year later, BCVA was found to be 20/29 in both eyes. Automated visual field testing showed marked improvement of the central scotoma and foveal threshold in both eyes, but a paracentral temporal visual field defect could still be found bilaterally (Fig. 1), while spectral-domain optical coherence tomography showed a diffuse loss of peripapillary RNFL thickness and macular ganglion cell layer thickness in both eyes.

Discussion

Variants of nucleotide m.13042 of the mitochondrial ND5 gene have been associated with a LHON-like phenotype in a handful of previous reports, and has currently been classified as “likely pathogenic” for mitochondrial disease by the ClinGen Mitochondrial Disease Nuclear and Mitochondrial Variant Curation Expert Panel [5, 8–10]. The same heteroplasmic transition has been identified in an Italian family, with documented LHON-like optic neuropathy in two individuals [8]. These two affected individuals, as well as some of the unaffected carriers, also presented with retinopathy (characterized by small yellowish dots distributed both in the peripheral and central retina) and early-onset cataract. Reconstruction of the family history might also suggest the possibility of mitochondrial encephalomyopathy, lactic acidosis, stroke-like syndrome (MELAS) and Leigh clinical phenotypes [8]. Additionally, a Polish report of a single patient with the same substitution described a progressive, predominantly neurological phenotype with overlapping symptomatology of mitochondrial encephalomyopathy, lactic acidosis, stroke-like syndrome (MELAS), myoclonic epilepsy, ragged red fibers (MERRF), progressive external ophthalmoplegia (PEO) and finally LHON [9]. An alternative homoplasmic gene missense variant m.13042G>T, resulting in a different amino acid change p.(Ala236Ser), has been identified in a Slovenian family, with documented bilateral optic neuropathy in three individuals, in the absence of extraocular clinical features [5].

Eye fundus examination in our patient revealed hyperemic optic discs, retinal vascular tortuosity and peripapillary telangiectasia, all reminiscent of LHON (Fig. 2). No associated cataract, central maculopathy, or peripheral retinopathy was observed during follow-up. No other neurologic symptoms or signs were present. The patient had minor complaints of a cough and a sore throat at initial presentation, with positive COVID-19 screening. COVID-19 vaccination and infection have been (cautiously) proposed as a possible trigger for LHON activation in carriers [11]. Our patient also presented with parotitis 3 weeks after initial presentation, but elevated CRP and white blood cell count (not present initially), and rapid clinical and biological response with antibiotics would indicate an unrelated infectious cause.

The peak age of onset in LHON is between 15 and 30 years, and visual loss is usually severe with visual acuity reduced to 20/200 or less, dense central or centrocecal scotoma, and profound impairment in color vision [2, 3, 6]. Overall, visual prognosis is poor with most patients showing no functional improvement [2, 3, 6]. However, spontaneous (partial) visual recovery can be observed, most likely within the first year, but it may also occur several years or even decades following the initial visual loss [2, 3, 6]. Visual recovery rates are influenced by the patient’s genotype: of the three primary pathogenic mtDNA alterations, the m.14484T>C alteration is believed to carry the best prognosis, while the m.11778G>A variant has been associated with the worst prognosis, and the m.3460G>A variant with an intermediate visual prognosis [2, 3, 6]. Early age of onset, subacute presentation with slow visual deterioration, and a large optic nerve head surface area have been reported as positive prognostic factors for visual recovery [2, 6]. In 2015, the European Medicines Agency (EMA) approved idebenone, a synthetic analogue of ubiquinone (also known as Coenzyme Q10), for treating adolescent and adult patients with LHON with a recommended dose of 900 mg/day (300 mg, 3 times a day). Treatment response has been reported to vary according to the underlying causative mtDNA mutation, disease phase, age, and gender [12]. The likelihood of clinically relevant recovery was found to increase with treatment duration, and a treatment duration up to 24 months has been recommended to maximize the probability of visual recovery [12].

Our patient experienced profound bilateral sequential visual loss at the age of 18, with BCVA eventually degrading to “counting fingers” in both eyes. However, progressive recovery of visual acuity was later observed in both eyes, resulting in BCVA of 20/29 in both eyes after 4 years of follow-up. This clinical improvement was most likely induced by a natural history of spontaneous recovery rather than the result of his short-term treatment (approximately 1 month) with idebenone. The onset of visual loss in both Italian patients with the same substitution as our patient was later (aged 39 and 35, respectively) but was also associated with good spontaneous visual recovery. Idebenone (270 mg/day increasing to 320 mg/day) was initiated in the proband at a later stage (more than 10 years after initial visual loss) but was discontinued after a year in absence of further therapeutic response [8]. Two of the Slovenian patients experienced profound visual acuity loss at an early age (20 and 11 years old, respectively), without visual recovery during follow-up (with the proband receiving idebenone 5 years after disease onset). The third Slovenian patient suffered from bilateral visual loss at the age of 58 years, with no follow-up data available [5]. The Polish report emphasizes on changing phenotypic expression in a patient with a predominantly neurologic clinical phenotype and lacks (detailed) information regarding visual function [9].

The three primary mutations are usually homoplasmic, but 10–15% of individuals who harbor a primary LHON mutation are thought to be heteroplasmic [1, 2, 4–7]. Heteroplasmy is thought to influence disease penetrance in LHON based on the observation that the risk of disease manifestation significantly reduced if the mutational load was less than 60%, the threshold required for triggering deleterious consequences of most mtDNA alterations on oxidative phosphorylation [1–3, 6]. However, heteroplasmic inheritance does not necessarily imply a more benign clinical course of the disease, and no correlation between the clinical manifestations of LHON and the percentage of mtDNA mutation was found in a study performed specifically in a LHON heteroplasmic population [1, 13]. Meanwhile, incomplete penetrance has also been observed among heteroplasmic carriers exceeding suprathreshold mutational levels [3, 6]. In a population-based study in the North East of England studying 16 families with a primary pathogenic LHON variant, approximately 33% of families had at least one heteroplasmic individual and none of the affected individuals harbored less than 40% variant mtDNA in their blood leukocytes [4].

DNA sequencing of peripheral blood leukocytes in our affected patient resulted in a variant load of 23.8%, considerably lower than 60% or even 40%, and even much lower mutational loads were found in both affected individuals in the previously mentioned Italian family [8]. In this family, strikingly lower mutational loads were found in blood (4% in the proband and 3% in the proband’s brother) compared with urinary epithelium (39% and 41%, respectively) and skeletal muscle (91% in the proband) [8]. Importantly, the Italian report provided evidence of the functional relevance of the m.13042G>A substitution as supported by the consistent increase of serum lactic acid, mitochondrial proliferation at muscle biopsy of the proband, and phosphorus magnetic resonance spectroscopy investigations, demonstrating a defect of oxidative metabolism in both brain and skeletal muscle of the proband. The authors therefore concluded that these findings conclusively established the pathogenic role of the m.13042G>A variant [8].

Interestingly, LHON clinical phenotypes associated with a relatively low variant load have been observed for other point mutations in the mitochondrial ND5 gene, including m.13094T>C and m.13513G>A mutations [14–16]. Limited improvement in visual function has been reported in such patients with a m.13513G>A mutation; while BCVA increased from “counting fingers” in both eyes to 20/200 in the right eye and to 20/320 in the left eye in a patient who was being treated with idebenone 900 mg per day, spontaneous improvement from “counting fingers” to 20/320 in the right eye and from “hand motion” to “counting fingers” in the left eye was found in another patient [15, 16]. Where available, discrepancies of heteroplasmic levels with consistently lower mutational loads in blood were found, compared to other tested tissues such as urine, muscle, and kidney [14, 16]. Previous studies have demonstrated that ND5 polypeptide synthesis is probably the rate‐limiting step for complex I activity [14, 15, 17]. Therefore, pathogenic variants in the mitochondrial ND5 gene may cause a significant complex I defect, resulting in severe phenotypic expression even when present at a low load [14, 15, 17].

Conclusion

This case is an important reminder that complete mitochondrial genome sequencing should be considered if there is a strong clinical suspicion of LHON and when initial molecular testing (typically limited to the three primary mutations) is negative. Furthermore, discrepancies of heteroplasmy levels between different tissues (with consistently lower blood loads) have been observed for several MT-ND5 point mutations that resulted in LHON-like clinical phenotypes. Therefore, testing other tissues than peripheral blood (typically urinary epithelium or skeletal muscle) might sometimes be required to obtain (or confirm) the correct diagnosis.

Acknowledgment

This case report does not include any non-author contributors to acknowledge.

Statement of Ethics

Written informed consent was obtained from the patients for publication of this case report and any accompanying images. Approval by the Ethics Committee of the CHU Saint-Pierre hospital in Brussels was acquired on July 7, 2025 (reference number CE/25-07–11). The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000551571).

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

T.B., A.M., and F.W. were involved in the diagnostic process and clinical follow-up of the patient, while S.S. offered expert opinion. T.B. wrote the original draft and conducted the relevant literature search associated with it. All authors read, reviewed, and approved the final manuscript.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.