8-year retrospective analysis of intravenous arginine therapy for acute metabolic strokes in pediatric mitochondrial disease

Abstract

Background: Intravenous (IV) arginine has been reported to ameliorate acute metabolic stroke symptoms in adult patients with Mitochondrial Encephalopathy with Lactic Acidosis and Stroke-like Episodes (MELAS) syndrome, where its therapeutic benefit is postulated to result from arginine acting as a nitric oxide donor to reverse vasospasm. Further, reduced plasma arginine may occur in mitochondrial disease since the biosynthesis of arginine's precursor, citrulline, requires ATP. Metabolic strokes occur across a wide array of primary mitochondrial diseases having diverse molecular etiologies that are likely to share similar pathophysiologic mechanisms. Therefore, IV arginine has been increasingly used for the acute clinical treatment of metabolic stroke across a broad mitochondrial disease population.

Methods: We performed retrospective analysis of a large cohort of subjects who were under 18 years of age at IRB #08-6177 study enrollment and had molecularly-confirmed primary mitochondrial disease (n = 71, excluding the common MELAS m.3243A>G mutation). 9 unrelated subjects in this cohort received acute arginine IV treatment for one or more stroke-like episodes (n = 17 total episodes) between 2009 and 2016 at the Children's Hospital of Philadelphia. Retrospectively reviewed data included subject genotype, clinical symptoms, age, arginine dosing, neuroimaging (if performed), prophylactic therapies, and adverse events.

Results: Genetic etiologies of subjects who presented with acute metabolic strokes included 4 mitochondrial DNA (mtDNA) pathogenic point mutations, 1 mtDNA deletion, and 4 nuclear gene disorders. Subject age ranged from 19 months to 23 years at the time of any metabolic stroke episode (median, 8 years). 3 subjects had recurrent stroke episodes. 70% of subjects were on prophylactic arginine or citrulline therapy at the time of a stroke-like episode. IV arginine was initiated on initial presentation in 65% of cases. IV arginine was given for 1-7 days (median, 1 day). A positive clinical response to IV arginine occurred in 47% of stroke-like episodes; an additional 6% of episodes showed clinical benefit from multiple simultaneous treatments that included arginine, confounding sole interpretation of arginine effect. All IV arginine-responsive stroke-like episodes (n = 8) received treatment immediately on presentation (p = .003). Interestingly, the presence of unilateral symptoms strongly predicted arginine response (p = .02, Chi-Square); however, almost all of these cases immediately received IV arginine, confounding interpretation of causality direction. Suggestive trends toward increased IV arginine response were seen in subjects with mtDNA relative to nDNA mutations and in older pediatric subjects, although statistical significance was not reached possibly due to small sample size. No adverse events, including hypotensive episodes, from IV arginine therapy were reported.

Conclusions: Single-center retrospective analysis suggests that IV arginine therapy yields significant therapeutic benefit with little risk in pediatric mitochondrial disease stroke subjects across a wide range of genetic etiologies beyond classical MELAS. Acute hemiplegic stroke, in particular, was highly responsive to IV arginine treatment. Prospective studies with consistent arginine dosing, and pre- and post-neuroimaging, will further inform the clinical utility of IV arginine therapy for acute metabolic stroke in pediatric mitochondrial disease.

Keywords: Brain disease; Inborn error; Leigh syndrome; Metabolic stroke; Mitochondrial encephalomyopathies; Treatment.

Figure 1. Response to IV arginine therapy at time of acute stroke-like symptoms by age

N/A = not clearly assessable due to multiple simultaneous treatment interventions at time of acute stroke-like symptoms. (p=0.24 by Mann-Whitney U Test.)

Figure 2. Response to IV arginine therapy by key factor analysis

IV arginine treatment responders and non-responders are shown on the right and left side of each graph, respectively. Bubble sizes correspond to subject number per group. A. Subjects are grouped by clinical response and whether IV arginine treatment was given immediately on hospital presentation (top) or not (bottom) (p= 0.003, Chi-Square). B. Subjects are grouped by clinical response and the presence (top) or absence (bottom) of hemiplegia as a presenting symptom (p=0.002, Chi-Square). C. Subjects are grouped by clinical response and molecular lesion: nuclear DNA (bottom) or mtDNA (top) (p=0.10, Chi-Square).

Figure 3. Brain neuroimaging studies in subjects who received IV arginine for acute clinical stroke-like episodes

T2-weighted Brain MRIs available before and after acute stroke-like episodes (SLE) for each study subject (diffusion-weighted images are not shown, but are described). A. Subject 1, brain MRI 2 years prior (left) and 4 weeks after (right) SLE & IV arginine treatment showed new T2 prolongation in periaqueductal gray, not accompanied by diffusion restriction. B. Subject 2, brain MRI 6 months prior (left) and 7 months after (right) SLE & IV arginine treatment showed new T2 prolongation in left greater than right posterior putamen, which was also associated with new T2 prolongation in the superior colliculi and left caudate (not shown); there was accompanying diffusion restriction. C. Subject 3, brain MRI 13 months prior to SLE (top first panel); immediately after first SLE (bottom first panel) shows new T2 prolongation in the thalamus with a mixed diffusion pattern (SLE was not treated and not described in the table as it occurred prior to diagnosis); 2 years and 1 month prior to 2^nd SLE (top 2^nd panel) and 1 day after second SLE; on day 2/7 of IV arginine, (bottom 2^nd panel) MRI showed new T2 prolongation in the cortical and subcortical areas in the left parietal and occipital regions more than the left frontal, right parasaggital parietal and left thalamus, all with diffusion restriction (not shown); 3 months prior to (top 3^rd panel) and 1 day after third SLE, on day 2/3 of IV arginine (bottom 3^rd panel) MRI showed new T2 prolongation in the right parasagittal frontal with diffusion restriction; 1 week before (top 4^th panel) and 1 day after fourth SLE & IV arginine treatment (bottom 4^th panel) showed new T2 prolongation in the left medial occipital gyrus with subtle diffusion restriction. D. Subject 4, brain MRI at first SLE (first panel) showed T2 hyperintensity in the medial right occipital lobe, with extension into the posterior temporal lobe, associated with restricted diffusion (this SLE was not treated and not described in the table as it occurred prior to diagnosis.); brain MRI 1 day after 2^nd SLE during day 1/1 of IV arginine infusion (2^nd panel) showed no new changes; brain MRI 1 day after 3^rd SLE & IV arginine treatment (3^rd panel) shows T2 hyperintensity in medial left occipital lobe associated with diffusion restriction, with increased perfusion through remainder of left occipital lobe; brain MRI during 4^th SLE, prior to IV arginine treatment (4^th panel) showed subtle T2 hyperintensity in the medial right parieto-occipital and occipital region, with diffusion restriction, as well as punctate areas of T2 hyperintensity with diffusion restriction in the bilateral pre- and postcentral gyri; brain MRI immediately after 7^th SLE, after arginine treatment 1/1 (5^th panel) showed no new changes. Subject 5, brain MRI 2 years prior (left) and 13 days after SLE (right), 1 day prior to IV arginine treatment showed new T2 hyperintensity in the cortex, predominately in the frontoparietal region, with diffusion restriction (changes were present, but very subtle immediately after SLE; images not shown.) F. Subject 7, brain MRI 1 day after SLE & IV arginine treatment showed no specific signal abnormalities; incidental note of abnormal cortical morphology. G. Subject 8, brain MRI 2 months prior to (left) and 3 days after SLE & IV arginine treatment showed no interval change; the substantial apparent difference in T2 signal abnormality throughout the bilateral basal ganglia (and periventricular region, not shown) was attributed technical differences; the baseline MRI was performed in a different institution.