Corticospinal tract damage in HHH syndrome: a metabolic cause of hereditary spastic paraplegia

Hyperornithinemia–hyperammonemia–homocitrullinuria (HHH) syndrome (OMIM # 238970), caused by mutations in ORNT1 (SLC25A15) gene, is a rare autosomal recessive disorder of the urea cycle [1, 2]. The metabolic triad of hyperammonemia, hyperornithinemia and homocitrullinuria establishes the diagnosis. Clinical manifestations are widely heterogeneous, ranging from neonatal life-threatening hyperammonemic crises to milder forms with onset at variable ages. Once the treatment is started, the clinical course is usually stable, but a progressive lower limb spasticity and a cerebellar dysfunction appear in most cases, leading to mild gait disturbance up to loss of deambulation [3‐7]. Intellectual disability of variable degree and, more rarely, focal dystonia and myoclonus may also occur [3‐6]. Pathophysiology of pyramidal and cerebellar dysfunction remains to be elucidated. To date, neurophysiological and neuroimaging studies have been conducted only in few patients. Somato-sensory evoked potentials (SSEPs) and peripheral nerve conduction velocity (NCV) were occasionally found to be abnormal [6]. Motor evoked potentials (MEPs) have been rarely performed and, despite pyramidal dysfunction is presumed to be progressive, no longitudinal studies have been reported so far [4, 5]. Similarly, available neuroimaging data are mostly based on anecdotal descriptions [5, 8‐12]. Herein we describe neurological, neurophysiological and neuroradiological findings, collected over a 15-year follow-up, in 9 HHH patients. We aimed to correlate long motor and sensitive tracts and peripheral nervous system involvement with neuroimaging and neuro-functional outcome data.

We report on 9 patients (5 female, 4 male; mean age at last examination 24.4 ± 16.3 years; range 7–52.6), with genetically confirmed HHH syndrome, regularly followed at the Division of Metabolism of the Bambino Gesù Children’s Hospital in Rome. The early course and molecular abnormalities of 8/9 patients have been reported in previous studies [3, 5, 12‐15]. Clinical features are summarized in Table 1. Follow-up had a mean duration of 15.9 ± 12.6 years (range 3–34 years) and encompassed clinical, neurological, neurophysiological and neuro-radiological data, retrospectively reviewed and prospectively integrated whenever necessary.

This study reports on the longitudinal description of neurological, neurophysiological and neuroradiological patterns in HHH syndrome. In our cohort, pyramidal signs were invariably present, with an overt picture of spastic paraplegia detectable in one third of the cases. Other clinical features encompassed a cognitive impairment of variable degree, a frequent cerebellar dysfunction, and more rarely a secondary peripheral neuropathy, which enlarges the spectrum of the phenotypic disease severity. Characteristically, clinical signs of central sensory dysfunction were never observed. Neurophysiological data confirmed a prominent involvement of the primary motor descending pathway compared to the long fibers sensory tracts, and a less frequent involvement of peripheral nerve. Neuroradiological studies, despite not completed by diffusion tensor imaging techniques, confirmed the prominent affection of the regions anatomically corresponding to pyramidal tract (frontal subcortical white matter and, in the most severe cases, cortical gray matter as well). With regard to the final clinical outcome, our findings show heterogeneous profiles, which are apparently independent from the age at evaluation, from the age of treatment initiation, as well as from the length of the disease course, which is particularly difficult to assess in patients who never experienced overt hyperammonemia, which correspond to one third of our cohort.

The association of lower limbs spasticity with mild cerebellar signs, cognitive impairment, and possible neuropathy indicates that HHH syndrome should be listed among the hereditary spastic paraplegias (HSPs). In particular, given its composite phenotypic picture, HHH syndrome fulfills the diagnostic criteria of complex HSPs [25].

Nowadays, HSPs encompass more than 80 loci and over 60 genes [26, 27], nevertheless, many conditions remain undiagnosed, suggesting that new genetic determinants still remains to be identified.

With regard to neurophysiological findings in HHH syndrome, our data are consistent with the clinical picture, highlighting the predominant involvement of the descending motor tracts compared to the ascending sensory ones. MEPs recorded at lower limbs showed delayed or not evocable response in about 90% of our patients. This occurred in both adults and children, showing a progressive worsening trend over time. A slowing-down, up to the loss, of an evocable response of CMCT was mostly detectable at rest at first evaluation, but it progressively involved the response also at muscle contraction at follow-up. This is in line with the neurophysiological response to a progressive neuronal degeneration in which a pathological response at rest can be hided or mitigated in facilitation conditions, when higher threshold and faster pyramidal neurons are recruited [28, 29]. What follows is that the SPRS scale, which scores spastic paraplegia according to patient’s age and disease duration, is not an optimal tool to quantify the severity of pyramidal dysfunction in HHH syndrome.

Contrarily to MEPs, SSEPs resulted unaltered in the most of our patients. Accordingly, clinical signs of sensory defects were never recorded, not even in the two patients who showed mildly delayed SSEPs values.

Neuroradiological data also confirm a slow progression of the disease. Atrophy primarily affects the subcortical white matter and, in the most severe cases, mildly involving cortical gray matter as well, whilst cerebellum is less affected. These findings are in line with literature reports on brain lesions observed in urea cycle disorders, which selectively and primarily affects the deep white matter [30]. Since the disease course in HHH syndrome is usually stable, with low risk of hyperammonemia once diet and pharmacological therapy are started, brain lesions are milder if compared to other urea cycle disorders, presenting a more severe long-term course [31, 32]. Therefore, brain abnormalities detected in HHH syndrome are likely to be not only dependent on the severity and duration of hyperammonemia alone. In fact, the 4/9 patients in our cohort who experienced hyperammonemic coma showed heterogeneous MRI findings: #3 and #6 had the most severe clinical and neuroradiological outcome, while #1 and #4 had mild lesions, comparable to those of patients who never experienced severe hyperammonemia.

Concerning spinal cord, MRI changes suggest a correlation with the clinical outcome, with the most severe atrophic degeneration observed in patients with the higher SPRS score (patient #3 who is wheelchair bound, and patient #2 who hardly maintains autonomous gait, only for a few steps). Although spinal cord atrophy progression rate cannot be determined since patients underwent a single MRI study, the ongoing trend of neurophysiological assessments suggest a similar evolution.

Clinical, neurophysiological and neuroradiological findings in HHH syndrome resemble those observed in argininemia and pyrroline-5-carboxylate synthetase (P5CS) deficiency [33‐36], two other disorders of aminoacid metabolism connected to the distal part of the urea cycle (Fig. 4).

It is noteworthy that, as in HHH syndrome, these diseases are associated with spastic paraplegia, variable cognitive impairment and possible cerebellar signs, then suggesting possible common mechanisms connecting urea cycle-related pathways with the maintenance of the cortico-spinal tract integrity [37]. Possible shared mechanisms include abnormalities of arginine, creatine, polyamines, and proline metabolism, and dysregulation of the autophagy machinery, this last representing a known cause of some HSPs [38].

Our study highlights the progressive involvement of the corticospinal tract in HHH syndrome. With regard to neurophysiological data, corticospinal system dysfunction resulted clearly predominant if compared to central sensory system and to peripheral nervous system. MEPs showed a pathological pattern with a progressive worsening trend with age. Although the cause of the selective involvement of cortical spinal tract in HHH syndrome remains to be elucidated, the similarities with argininemia and P5CS deficiency, suggest possible common pathophysiological mechanisms.

The presence of pyramidal signs/lower limb spasticity associated with cerebellar signs and cognitive impairment, highlights the importance to list HHH syndrome among the complex form of HSPs.