Background
The antiepileptic drug sulthiame (STM),
N-(4-sulfamoylphenyl)-1,4-butansultam, was developed in the 1950s and introduced in the market in the 1960s. STM proved to be effective especially in focal seizures, particularly in benign epilepsy of childhood with centrotemporal spikes (BECTS) [
1‐
4]. Moreover, several open prospective and retrospective studies showed efficacy of STM as add-on treatment in cryptogenic or symptomatic localization-related epilepsies in children and adults [
2]. For treatment of BECTS and related epileptic syndromes a dose of 4 to 8 mg/kg body weight (BW) per day is recommended, and higher doses in West syndrome.
STM is available in Argentina, Australia, Czech Republic, Germany, Greece, Latvia, Hungary, Israel, Slovakia, and Switzerland [
2]. It is widely used in genetic, cryptogenic, and symptomatic focal epilepsies, especially in childhood and adolescence [
2].
Several studies demonstrated a linear correlation between the STM dose/BW and STM serum concentrations in children and adults [
1,
4].
STM is a sulfonamide derivate [
4]. Alike the antiepileptic drugs zonisamide and topiramate, STM is an inhibitor of mammalian carbonic anhydrase isoforms I-XIV [
5]. Carbonic anhydrase inhibitors may cause bone marrow depression, skin toxicity, sulfonamide-like renal lesions, and allergic reactions in hypersensitive individuals. Dependent on the dosage patients may show drowsiness and paresthesias. Largely, adverse effects are secondary to urinary alkalinization or metabolic acidosis [
6,
7].
Well-known adverse effects of STM comprise paraesthesias of face and limbs, exercise hyperpnoea, dyspnoea, and anorexia. Rare side effects include cognitive impairment, elevation of liver enzymes, acute renal failure, and metabolic acidosis [
8‐
13]. Visual loss or any optic nerve damage were never reported as a side effect of STM.
Here, we report two unrelated children carrying two different mutations of mitochondrial DNA associated with Leber hereditary optic neuropathy (LHON), who suffered visual loss in close temporal connection with STM medication for focal epilepsy. We discuss whether this is pure coincidence or reflects STM being a trigger factor for onset of visual failure in LHON mutation carriers. Based on biochemical investigations in fibroblasts we provide evidence for marked impairment of oxidative phosphorylation caused by STM.
Discussion
We describe two unrelated patients, an 8-year-old girl (subject #1) and an 11-year-old boy (subject #2), who suffered acute visual loss in close chronological connection with starting STM treatment for cryptogenic focal epilepsy. Subject #1 had already a mild bilateral visual impairment and an established LHON diagnosis (associated with the m.14484T>C mutation) before initiation of STM treatment. Her visual acuity markedly decreased within a few days after starting of STM treatment, and worsened further within a few days of increasing the STM dosage.
Subject #2 had acute monocular painless visual loss the day after increasing the STM dosage, at age 11 years 4 months. He was later diagnosed with LHON associated with the m.3460G>A mutation. In both subjects, there was no evidence for any other environmental factors known to trigger visual loss in LHON mutation carriers.
The question of whether these observations reflect a causal relationship between STM intake and triggering the onset of LHON-associated visual loss, or whether this is mere temporal coincidence, is difficult to assess.
LHON (MIM #535000), the most common mitochondrial disorder, is a maternally inherited disease characterized by acute or subacute painless visual loss [
14‐
17]. Visual failure occurs most often in young adult life, with a clear predominance of affected males (5:1 male to female ratio) [
17]. In the acute phase, patients describe loss of colour vision and visual blurring in the central visual field in one eye. Visual field testing shows a progressively enlarging centrocecal scotoma, and in most cases, there is severe visual loss, with markedly reduced VA. Usually within 2 or 3 months, the second eye is similarly affected, but bilateral onset of visual loss occurs in about 25% of cases. Later optic atrophy develops, with lifelong bilateral visual loss, leaving most patients legally blind [
15].
The vast majority of European and North American patients carry one of three pathogenic mutations of the mitochondrial DNA (m.11778G>A, m.3460G>A, m.14484T>C), which lead to dysfunctional complex I of the mitochondrial respiratory chain [
15]. This results in a defect of adenosine triphosphate (ATP) synthesis and an enhanced production of reactive oxygen species (ROS), leading to dysfunction and loss of a proportion of retinal ganglion cells [
16].
Genotype–phenotype correlations have been recognized, and the m.14484T>C variant in the MT-
ND6 gene is related to a most favorable long-term visual outcome [
15].
Penetrance of LHON is incomplete, visual loss arises in only ca. 50% of males and ca. 10% of females carrying one of the three primary LHON mutations. Occasional discordance of LHON in monozygotic twins indicates that environmental factors may have an impact on penetrance [
18]. Factors triggering visual loss in LHON mutation carriers include smoking, heavy intake of alcohol, and raised intraocular pressure [
19‐
21]. Poor nutrition, psychological stress, physical trauma, and acute illness have been blaimed as well [
18,
19].
The lifetime risk for visual failure in individuals carrying the homoplasmic mutation m.14484T>C of the mitochondrial DNA was estimated to be 47% in males and 8% in females [
22]. The mean age at onset of visual loss in females was 17.7 years, with a range of 11 to 21 years, in one case series [
22]. The onset of visual impairment already at age 7 years observed in subject #1—more than 1 year before STM treatment was started—is earlier than what would be expected in the literature. But for this patient a critical worsening of the visual loss occurred in close temporal relation with starting treatment with STM and increasing the dose of STM.
The lifetime risk for visual failure in individuals with the homoplasmic mitochondrial DNA pathogenic variant m.3460G>A was in one study reported to be 32% in males and 15% in females with a median age of onset of 20 years in males [
15]. A second study described a risk of 49% in males and 28% in females, with a median age of onset of 22 years (standard deviation 13.6 years) in males [
17]. The range of age of onset was not provided in these studies. Subject #2 suffered visual loss at age 11 years 4 months, relatively early in life. As for subject #1, visual loss occurred in close temporal relation with a dose increase of STM treatment.
Apart from visual impairment, most LHON patients are otherwise healthy. However, occasional associations with neurological, cardiac, and skeletal features were reported in some patients with LHON. Neurologic features described to occur in LHON more commonly than in the general population include postural tremor, peripheral neuropathy, nonspecific myopathy, and movement disorders. A multiple sclerosis-like condition (or “Harding´s disease”) was observed in some subjects with LHON, mostly women [
15,
23]. A recent nationwide study from Denmark revealed an increased incidence for stroke, demyelinating disorders, dementia, and epilepsy in individuals with LHON, but not in their family members [
24]. Focal epilepsy in childhood, as observed in the two subjects reported here, is unusual to occur in patients with LHON. We can only speculate whether LHON and epilepsy in these two subjects are caused by a unique pathomechanism or whether they represent two independent conditions.
A deleterious effect on mitochondrial function has been shown, or is pathophysiologically possible, for a wide range of medications [
25,
26]. Several substances known to interfere with the function of the mitochondrial respiratory chain are under suspicion of triggering LHON in mutation carriers. These include antibiotics such as ethambutol, chloramphenicol, linezolid, aminoglycosides; antiretroviral drugs; as well as cyanides, methanol, pesticides, and phosphodiesterase type 5 inhibitors [
25‐
30].
Another mitochondrial carbonic anhydrase inhibitor used for antiepileptic treatment, topiramate (TPM), was reported to have triggered visual loss in an adult patient carrying the LHON m.11778G>A mutation, who had received TPM for temporal lobe epilepsy [
31].
The pathomechanisms of triggering of onset of visual loss onset in LHON mutation carriers by carbonic anhydrase inhibitors is difficult to elucidate. Previously reported animal experiments point to an impairment of oxidative phosphorylation caused by metabolic acidosis [
32‐
34]. Thus, the metabolic acidosis known to occur under medication with carbonic anhydrase inhibitors is a possible explanation leading to impairment of mitochondrial function, resulting in visual loss in LHON mutation carriers. However, while an impact of STM-associated metabolic acidosis on mitochondrial function seems theoretically possible, reports of significant acidosis under STM treatment are rare [
8,
12], and it was shown that urinary pH remains in the normal range under STM treatment [
35]. Furthermore, in this study we did not observe a significant increase of extracellular acidification rates in LHON fibroblasts compared to control cells when treated with STM. Thus, STM-associated metabolic acidosis may possibly play only a minor role in triggering onset of visual failure in LHON.
Mitochondria are impermeable to HCO3
−. The two mitochondrial carbonic anhydrase isoforms VA (CA-VA) and VB (CA-VB) are responsible for providing HCO3
− to essential mitochondrial enzymes [Carbamoyl-phosphate synthase (CPS1), pyruvate carboxylase (PC), propionyl-CoA carboxylase (PCC), 3-methylcrotonyl-CoA carboxylase (3MCC)]. These enzymes catalyze the production of metabolic intermediates of the urea and Krebs cycle [
36]. Thus, it is tempting to speculate that STM diminishes the production of reducing equivalents through the Krebs cycle, decreases ATP production and impairs oxidative phosphorylation. However, at this point we can not exclude that STM could also affect functionality of the OXPHOS system in a more direct manner. Further analyses will be required to address this question.
In fact, we provide evidence from real-time respirometry in LHON fibroblasts that STM drastically deteriorates their respiration rate when treated with STM compared to vehicle control. We conclude that STM has a deleterious effect on cells with an already compromised mitochondrial function, which may explain the onset or worsening of visual loss under STM treatment.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.