A Proposed Diagnostic Algorithm for Inborn Errors of Metabolism Presenting With Movements Disorders

Inherited metabolic diseases or inborn errors of metabolism frequently manifest with both hyperkinetic (dystonia, chorea, myoclonus, ataxia, tremor, etc.) and hypokinetic (rigid-akinetic syndrome) movement disorders. The diagnosis of these diseases is in many cases difficult, because the same movement disorder can be caused by several diseases. Through a literature review, two hundred and thirty one inborn errors of metabolism presenting with movement disorders have been identified. Fifty-one percent of these diseases exhibits two or more movement disorders, of which ataxia and dystonia are the most frequent. Taking into account the wide range of these disorders, a methodical evaluation system needs to be stablished. This work proposes a six-step diagnostic algorithm for the identification of inborn errors of metabolism presenting with movement disorders comprising red flags, characterization of the movement disorders phenotype (type of movement disorder, age and nature of onset, distribution and temporal pattern) and other neurological and non-neurological signs, minimal biochemical investigation to diagnose treatable diseases, radiological patterns, genetic testing and ultimately, symptomatic, and disease-specific treatment. As a strong action, it is emphasized not to miss any treatable inborn error of metabolism through the algorithm.


INTRODUCTION
Inborn errors of metabolism (IEMs) are defined as any condition that leads to a disturbance of a metabolic pathway, irrespective of whether it is associated with abnormalities in biochemical laboratory tests. Indeed, IEMs include not only enzymes or transporters deficiencies or superactivities, but also chaperons and transcription factors abnormalities. Recently, a current nosology by Ferreira et al., defines 1,015 IEMs in 130 groups (1).
IEMs encompass a large group of single gene disorders that can affect all organs and lead to a variety of symptoms (2). Most of the IEMs are multisystem diseases with neurological (intellectual disability/developmental delay, cognitive regression, hypotonia, spasticity, neuropathy, vision and hearing impairment, encephalopathy/stroke, epilepsy, etc.) and non-neurological (failure to thrive, vomiting, hepatomegaly, splenomegaly, renal tubular acidosis, nephrolithiasis, etc.) manifestations. Commonly initial symptoms and signs of IEMs are somewhat nonspecific (3). Movement disorders (MD) are among the most usual neurological symptoms in children with IEMs and account for a significant part of the morbidity and mortality. MD manifest in IEMs that cause diffuse CNS or selective basal ganglia involvement. As it is known, the basal ganglia participate in the control of voluntary movement and on the other hand this brain area is especially vulnerable to certain IEMs as metal storage defects, energy metabolism and lysosomal storages disorders (4,5).
A central problem is the large number and variety of IEMs with MD and the poor recognition of these disorders that make it challenging for the pediatric neurologist to decide on the initial evaluation with the consequent delay in diagnosis and timely treatment. This work herein facilitates an update information with new IEMs that present MD, suggests red flags and diagnostics clues for suspecting IEMs, proposes the minimum biochemical studies as stated in each MD and the differential diagnoses according to the neuroradiological findings and provides evidence on symptomatic or disease specifictreatment through a six-step algorithm.

OVERVIEW OF MOVEMENT DISORDERS IN INBORN ERRORS OF METABOLISM
Based on the distinct pathway involved, IEMs can be categorized as described by the nosology of Ferreira et al. into the following groups (1): (1) Disorders of nitrogen-containing compounds, (2) Disorders of vitamins, cofactors, metals, and minerals, (3) Disorders of carbohydrates, (4) Mitochondrial disorders of energy metabolism, (5) Disorders of lipids, (6) Disorders of Tetrapyrroles, (7) Storage disorders, (8) Disorders of peroxisome and oxalate, and (9) Congenital disorders of glycosylation. This accurate and updated nosology will be applied in this review, in which a total of 231 IEMs presenting with MD will be included. The search strategy, the selection criteria and the data extraction applied are described in Supplementary File 1. The IEMs presenting with MD appear in Supplementary Table 1.
Whereas, MD are classified into two main categories, hyperkinetic, and hypokinetic movements. Hyperkinetic MD are unwanted or excess movements which include dystonia, choreoathetosis, tremor, myoclonus, tics, and stereotypies. Definition and classification of pediatric hyperkinetic movements were described by Sanger et al. (6), although; this classification did not include ataxia. Hypokinetic movements are called hypokinetic-rigid syndrome or Parkinsonism (7). Both hyperkinetic and hypokinetic MD are crucial clinical findings with significant implications for diagnosis and treatment.
MD in the context of IEMs are generally a combination of different MD, therefore a MD oftentimes does not predict the type of IEM. The 51% of IEMs of this review presents 2 or more MD. In this group there are well-known diseases namely neurotransmitters disorders (8), cerebrotendinous xanthomatosis (9), Wilson disease (10), POLG deficiency (11), Niemann Pick disease type C (12), and Glucose transporter 1 deficiency (13) that can exhibit virtually any MD and that must always be incorporated in the differential diagnosis on either child with MD. It should be noted that many of these referred IEMs are also treatable.
Several studies have shown that IEMs account for less than a quarter of diagnoses in children with MD, for this reason timely recognition of these underlying IEMs often allows disease-specific treatments and the best possible outcome for the affected child (3,28). The percentage of diagnosis reached depends on the population included and on the biochemical or genetics analysis conducted. Diagnosis rate is lower in children with acute symptoms and substantially higher in children with chronic pathology. In this review, six studies investigating IEMs presenting with MD in 909 children and adults have been found (3,(29)(30)(31)(32)(33). The groups examined were very heterogeneous while other study exclusively included patients with dystonia (29). When all movement disorders were included, dystonia and ataxia were the most frequent MD. Most of the studies applied targeted next generation sequencing panels obtaining a diagnostic yield ranged between 11 and 51%. Nevertheless, among three other studies in children with acute pathology presenting to the pediatric emergency department, the percentage of children with IEMs was between two and eight percent of all cases. These cases included IEMs as various as ceroidolipofuscinosis, mucopolisacaridosis, pantothenate kinaseassociated neurodegeneration, glutaric aciduria type 1, MELAS, and Leigh syndrome (34)(35)(36). The full-description of all these studies is displayed in Supplementary Table 2.
Regrettably, there are only a few studies that make a comprehensive description of the MD in the IEMs and most of the knowledge is based on individual cases or series of very few patients. Quite apart from the fact that many of these report do not provide in-depth characterization of the MD. Some studies that include large series of patients with IEMs and MD are: Tetrahydrobiopterin (BH4) deficiencies (37), cobalamin-related remethylation disorders (38), SERAC1 deficiency (39), mitochondrial disorders (22), PLA2G6associated neurodegeneration (24), lysosomal storage disorders (40), cerebrotendinous xanthomatosis (9), and congenital disorders of glycosylation (41). IEMs with MD as a primary or prominent feature and treatable causes should always be in mind. These groups of IEMs are listed in Table 1.
IEMs should repeatedly be considered, even in patients in whom an acquired cause of MD is assumed (e.g., infantile cerebral palsy). In a review by Leach et al. they identified 67 treatable IEMs that mimicked infantile cerebral palsy. Seventyfour percent of these IEMs have a specific treatment or a treatment that allows stabilizing/preventative effects. Fiftyseven percent of these IEMs are detected with straightforward metabolic/biochemical tests (80).
With regards to previous diagnostic algorithms of IEMs and MD, a simple scheme for adult patients was proposed by Sedel et al. that demonstrated a diagnostic approach based first on the clinical course of symptoms and the brain MRI (4). This very simple scheme allows the exclusion of the most frequent etiologies speedily whilst it cannot be applied altogether to childhood-onset MD because the diseases differ from adults.
One as well the other, IEM and MD impact intensively children's quality of life (QOL). In the study by Eggink et al., it was found that increasing severity and lower adaptative abilities of MD positively correlated with lower (QOL) scores (2). For this reason, adequate treatment can better ameliorate the quality of life of these children.

APPROACH TO CHILD WITH SUSPECTED IEM AND MD
In this work, a six-step diagnostic algorithm was drafted and represented in Figure 2. The algorithm is based on simple steps that are listed below. The algorithm helps decide whether to consider a metabolic cause in a patient with MD. It also guides in the characterization of the MD, considering the characteristics that will most help guide a defined IEMs. It presents a list of neurological and non-neurological symptoms to take into account to establish the differential diagnosis of IEMs. Dependent on each MD, it proposes a list of minimal biochemical studies and also a list of IEMs to consider in fulfillment of the neuroradiological findings. Finally, a summary of the available treatments of both the IEMs and the MD is made. Once more, it is very important to note that the algorithm encourages thinking about treatable causes across different steps of the process.

STEP 1. WHEN TO SUSPECT AN IEM IN MD?
IEMs should be included in the differential diagnosis of all children with MD, even though there are several signs and       2) Combination of different MD (ataxia plus dystonia, dystonia plus parkinsonism, chorea plus dystonia and myoclonus with any other MD or neurological or non-neurological signs, other than seizures), 3) Acute or subacute onset, remarkably if the onset is associated with encephalopathy/coma or if onset is precipitated by a concurrent febrile illness, starvation, physical exhaustion, or after a high-protein ingestion, 4) Insidious onset in a patient with multiple previous nonsystemic manifestations, 5) Consanguinity and/or recessive or X-linked inheritance pattern, 6) Distinct neuroradiological findings for instance basal ganglia abnormalities, white matter involvement (hypomyelination/ leukoencephalopathy) or cerebellar atrophy, 7) Atypical or progressively abnormal MD that fails to respond to standard treatment, and 8) MD that are not explained by classical etiologies (e.g., structural brain lesion, infectious/parainfectious or autoimmune disorders, toxic or drug induced MD and other well-known genetic or neurodegenerative MD, etc.).

Phenomenology of MD
Not surprisingly this is the most significant step in the algorithm and also the one that can offer the most difficulty. The characterization of the MD must be in line with the definitions of Sanger et al., and reinforce in the own experience of the physician (6). As far as possible it is of the utmost importance to have also the assessment of an expert in pediatric MD. In a straightforward manner, it should be determined: 1. If the child presents unwanted excess movement (hyperkinetic movements) will be classified taking into account the definitions for dystonia, chorea, myoclonus, tremor, and stereotypies fairly detailed by Sanger et al. (6). The definition of ataxia is not included in the latter; accordingly, this definition will be used: a. Ataxia: consists of the decomposition of movement due to breakdown of normal coordinated execution of a voluntary movement. Dysmetria, dysdiadochokinesia, intentional tremor, hypotonia, wide-based gait, or inability to tandem walk are frequently associated signs of ataxia. Early onset cerebellar ataxia consists in a large group of patients with rare disorders, manifesting symptomatic cerebellar ataxia before the age of 25 years (82).
It is worth noting that tics are not habitually associated with IEMs, hence have not been include in this article, despite the fact that few cases of patients with tics have been reported in the following IEMs: Lesch-Nyhan disease (83), juvenile neuronal ceroid-lipofuscinosis (84), pantothenate kinaseassociated neurodegeneration (85), Wilson disease (86), POLG disease (87), mannosidosis (88), COASY-associated neurodegeneration (89), and Maple Syrup Urine Disease (90). 2. If the child presents decreased (hypokinesia) and slow (bradykinesia) movements, rigidity or rest tremor. These are all signs of parkinsonism or hypokinetic rigid syndrome. These signs commonly do not appear all at the same time in a child, therefore the syndrome is routinely "incomplete" (7). The diagnosis of HRS greatly narrow possible IEMs. In this instance, is extremely helpful to observe the recommendations made by Garcia-Cazorla et al. to separate etiologies on the basis of age at onset (7). HRS starting at very young age (<2 years) can guide us to the diagnosis of monoamine neurotransmitter defects, mitochondrial diseases or Neurodegeneration with Brain Iron Accumulation disorders. On the contrary, HRS starting in children older than 2 year of age, can guide us in addition to those formerly mentioned in the following disorders: Wilson disease (10), GLUT-1 deficiency (13), Niemann-Pick type C (12), Gaucher disease (27), manganese accumulation disorders (16,56), glutaric aciduria type 1 (52), Celia's encephalopathy (91), gangliosidosis (26), and cerebrotendinous xanthomatosis (9). Although it should be noted that HRS generally appears predominantly in the initial stages of the monoamine neurotransmitter defects or mitochondrial disorders, emerging later in the remaining IEMs.
By virtue of the response to treatment, it is important to highlight a very important group of treatable IEMs frequently presenting with HRS: the monoamine neurotransmitter disorders.
The clinical phenotypes of these IEMs are predominantly neurological and the symptoms are similar to other childhood neurological disorders, such as cerebral palsy, hypoxic ischemic encephalopathy, or epileptic encephalopathies, as a result the monoamine neurotransmitter disorders are not recognized and are often misdiagnosed. Marked diurnal variation of motor symptoms (axial hypotonia and gait disturbances) and MD (dystonia, HRS, dyskinesia/chorea, tremor, and less commonly myoclonus) are often prominent along with autonomic dysfunction (sweating, hypersalivation, nasal congestion, and temperature dysregulation). Detailed MD phenotype, other clinical features as well as biochemical findings and neuroimaging of monoamine neurotransmitter disorders appears in Table 1. Prompt diagnosis is fundamental, considering that many respond to treatment and that adequate therapy is curative in some disorders. Conversely, the remarkable response to L-dopa of any childhood-onset undiagnosed MD should also guide us to this group of IEMs (8). Commonly a child affected by an IEM present more than one type of MD. To follow the proposed six-step algorithm, we should take into account all the MD that we can identify regardless the fact that we will always take into more consideration: (1) the most prominent MD in the patient or (2) the least frequent (more specific) MD in all cases. In this sense, diseases that show myoclonus or HRS are less numerous, limiting for the best our diagnostic options. A list of treatable IEMs according to the MD they present appears in Supplementary Table 3.

MD Age at Onset and Type of Onset
In some cases, the MD begins earlier in the prenatal or the neonatal period. With respect to infants and children, there are plenty IEMs presenting with MD that can debut at these group To be noted that many IEMs presenting with MD do not have a specific age of onset, hence usually this parameter may not be extremely valuable in evaluating a particular case.
of ages, consequently the isolated age of onset criteria will not be useful to differentiate between the different IEMs. Although, there are some IEMs presenting MD frequently before 2 years of age. Many MD in IEMs have an onset in adulthood regardless numerous cases with non-specific non-neurological symptoms as early as childhood or adolescence ( Table 2). MD that appear acutely, associated or not with encephalopathy, are generally caused by diseases that are considered medical emergencies. In cases where the onset of MD coincides with a metabolic decompensation, there is several times an acute injury to the basal ganglia or the cerebral cortex, in most cases with accompanying encephalopathy/coma, as is often the case in Leigh syndrome (22), glutaric aciduria type 1 (52), propionic acidemia (99), and other organic acidurias or mitochondrial disorders. Injury to the basal ganglia can also occur with the appearance of MD delayed as in the β-ketothiolase deficiency (111). In the remaining IEMs, there is very often no clear onset of MD disorder.
Notwithstanding, status dystonicus is a rare and lifethreatening MD emergency. It has been described in the following IEMs: Wilson disease (116), Pantothenate kinaseassociated neurodegeneration (117)(118)(119)(120), aromatic l-amino acid decarboxylase (AADC) deficiency (45) and SLC6A3 dopamine transporter deficiency syndrome (121). Management is adjusted to the characteristics and complications of each patient. The use of dystonia action plans can facilitate intervention or avoid progression to dystonic state. One of these plans have been describe in detail by Allen et al. (122). If the dystonic refractory state persists despite drug therapy, early consideration of intrathecal baclofen or deep brain stimulation is required.
Finally, involuntary MD can in addition appear during treatment, for example tremor with the administration of vitamin B12.

MD Distribution
Focal MDs in IEMs are rare and are generally associated with asymmetric radiologic injury. Multifocal or segmental MDs have been described in: dystonia: PMM2-CDG (41), myoclonus: Pyridoxamine 5 ′ -phosphate oxidase deficiency (123) and hemiplegic episodes: episodic ataxia type 6 (113) and MELAS (124). In the remaining IEMs, the distribution of the MD is usually generalized. In this context, the MD distribution is not very useful to guide the diagnosis.

MD Temporal Pattern
In the majority of IEMs, as expected, MD presented persistently and continuously, due to this circumstances, the temporary pattern does not usually provide diagnostic benefits. In contrast, it is very useful in a restricted group of IEMs presenting with paroxysmal or intermittent MD ( Table 3) as well as the monoamine neurotransmitter disorders whereby there is a very characteristic diurnal fluctuation of symptoms: motor symptoms, including MD, become more prominent in the evening and improve after sleep.

Other Neurological Signs
In the second step, the neurological evaluation is completed with other signs and symptoms that can guide the diagnosis. The following signs and symptoms will be taken into account: It is imperative to emphasize that the ophthalmological exploration can find numerous signs, which are ordinarily more or less specific and can exceedingly help guide the diagnosis of the child.

Non-neurological Signs
Thus, far as possible these non-neurological signs and symptoms will be considered during the evaluation of a patient with IEMs:

STEP 3. BIOCHEMICAL INVESTIGATION AND NEUROIMAGING
Eighty-eight percent of the 231 IEMs with MD can be detected through biochemical tests in blood, urine, feces, fibroblasts or CSF. Nevertheless, the diagnosis in the remaining IEMs will only be accomplish with the genetic testing. The tests to request are heterogeneous and will vary in line with the assumed MD and the interest in excluding certain diseases. First-tier and second-tier testing have been considered for each MD in order to prioritize the diagnosis of treatable diseases (see Table 4).
In the second-tier testing, more specialized or less common biochemical analysis (enzymatic activities, porphyrins, etc.) have been included. However, some tests of the second-tier should be considered at the beginning of the diagnostic process if the suspicion of a particular disease is very high. As expected, this list of biochemical tests is broader for patients with ataxia, the most frequent MD in patients with IEMs and the one with the most possible etiologies whereas the list become more restricted for patients with myoclonus and hypokinetic-rigid syndrome. Subsequent confirmation of a definite disease may be done through additional biochemical or genetic tests. In many countries where there is difficulty in performing biochemical tests, genetic testing should be done sooner rather than later. A
vital aspect in the biochemical study of a patient is to know at all times the diseases that have been ruled out with the requested tests and, therefore, to make a more thought-out plan for each patient. An alternative and sometimes controversial approach is the early performance of the lumbar puncture in a patient with MD. Lumbar puncture is regularly included in the initial evaluation of all children with MD, first and foremost when there is no speedy availability of genetic tests. This approach is correct unless there is a differential diagnostic hypothesis in which a lumbar puncture is not considered necessary. Thus, analysis of cerebrospinal fluid in IEMs presenting with MD is essential and should be considered as a 1st Tier biochemical investigation, especially in cases of ataxia, dystonia, or HRS, taking into account the need for the precise diagnosis of certain IEMs (monoamine neurotransmitter disorders, GLUT-1 deficiency, etc.) with the consequent establishment of a specific treatment (8,13). In this section it is necessary to comment on newborn screening programs since many patients are diagnosed now even before presenting neurological or non-neurological symptoms. The main purpose of newborn screening programs is to diagnose genetic disorders early, allowing treatment to begin before symptoms appear. The quantification of amino acids and acylcarnitines in dried blood spots (DBS) by MS/MS allows the simultaneous detection of more than 30 metabolic disorders, including those associated with amino acid, organic acid, and fatty acid metabolism. Other disorders detected using different techniques, e.g., galactosemia, biotinidase deficiency, Pompe disease, and mucopolysaccharidosis type I, are now included in some NBS programs. In a study by Navarrete et al., diagnosis of an inborn error of metabolism were confirmed in fifty-nine percent of patients tested using a personalized exome sequencing panel (142). Along with these baseline biochemical testing, neuroimage can identify patterns of treatable inborn errors of metabolism and helps to exclude other secondary causes of MD guiding further testing. Consequently, neuroimage should be considered part of the initial diagnostic evaluation (81). In chronic MD it is essential to make a correct assessment of the neuroimaging, which generally gives many clues to the specific diagnosis. In the IEMs that cause MD, a combination of cerebral atrophy, cerebellar atrophy, white matter alteration (hypomyelination, demyelination, nonspecific lesions, or decreased volume) and basal ganglia involvement (hyperintensity, metal deposit, etc.) can be observed. Additionally, less frequently injuries in other areas like the brain stem, corpus callosum or the spinal cord or metabolic can be detected.
In the study by Cordeiro et al., basal ganglia or gray matter abnormalities with or without white matter changes were present in sixty seven percent of the patients with IEMs (3). It is of the utmost importance to keep in mind that many IEMs can present normal neuroimaging in the initial stages and exclusively progress to cerebral or cerebellar atrophy in advanced stages of the disease. Therefore, it is necessary to be careful with the interpretation of a normal brain MRI. Spectroscopy can be useful in some cases: Creatine transporter deficiency (43), SLC13A5-Epileptic encephalopathy, early infantile (143), 25, Pyruvate dehydrogenase complex deficiency (58)(59)(60)(61), thiamine pyrophosphokinase deficiency, biotin-thiamine-responsive basal ganglia disease, mitochondrial thiamine pyrophosphate transporter deficiency (54), and many other mitochondrial disorders (Figure 4). Detection of intracerebral or basal ganglia calcification is also suggestive of certain IEMs (Table 5, Figure 3): Dihydropteridine reductase deficiency (37), folate receptor alpha deficiency, dihydrofolate reductase deficiency, hereditary folate malabsorption (144), phenylketonuria (144), Krabbe disease (145) and Aicardi-Goutières syndrome (146). Figure 3 and Figure 4 present the radiological pattern and MR Spectroscopy of some IEMs and Table 5 shows a list of IEMs with disease-specific treatment classified according to the neuroradiological pattern.

STEP 4. GENETIC INVESTIGATION IN IEMs
Genetic analysis is currently performed as a definitive test to confirm a diagnosis and to identify the exact gene involved. The  genetic testing strategy will depend on the clinical setting and the available resources. Diagnostic yield of karyotype, chromosomal microarray or single gene sequencing (unless there is a very clear biochemical biomarker) is in many cases very low and therefore should not be indicated as the first option. In the case of IEMs and MD, the recommended testing would be targeted next generation sequencing panels or whole exome/genome sequencing (81). The diagnostic yield of the different reported studies ranged between eleven and fifty one percent (see Supplementary Table 2). In a study by Cordeiro et al., targeted next-generation sequencing panels and whole exome sequencing increased diagnostic rate to more than 40%. In the same study, a genetic diagnosis provided either disease-specific or symptomatic therapy in thirty eight percent of the patients with a genetic diagnosis, highlighting the importance of genetic investigations to confirm underlying genetic cause in patients with pediatric MD. Early detection of treatable IEM allows for referral to expert centers and timely interventions to prevent disease progression and to improve neurological functioning. Additionally, detection of IEM for which no treatment currently exists allows for genetic counseling of the affected families (3).

STEP 5. TREATMENT OF PEDIATRIC METABOLIC MOVEMENT DISORDERS
MDs are a major medical problem. Treatment of children with IEMs-associated MD may include both symptomatic treatment of MDs and disease-specific management of the underlying IEM. Pharmacologic treatments for IEMs may comprise specialized dietary modifications, cofactor, vitamins, or supplements addition, enzyme replacement therapy, hematopoietic stem cell transplantation, and gene therapy (81). Successful treatment and better outcome frequent take place with early recognition of an underlying treatable IEM, therefore there should be an increased consciousness of these IEMs (147). A complete list of the specific treatments of each IEMs appears in Supplementary Table 1.
Regrettably, only thirty-eight percent of the IEMs included in this article have a disease-specific treatment. Although, over the last decades, new treatment approaches have changed the scope of IEMs from a group of rare, untreatable, and often fatal disorders to an important cause of potentially treatable diseases.
By contrast, when disease modification therapies are not possible, symptomatic therapy may be the only treatment that is available. Symptomatic therapy of MD requires multidisciplinary approach ranging from avoiding identifiable triggers, pharmacologic interventions, physiotherapy, to intrathecal baclofen pump, deep brain stimulation and gene therapy (3). Koy et al. has reviewed prodigiously the management of MD including dose range and guidelines for drugs commonly used to treat pediatric MD (148). The reader may refer to this review for more information. Even with treatment, MD hardly resolve completely and can cause life-long disability. Nevertheless, symptomatic treatment may improve the functional abilities and quality of life, correct orthopedic complications or relieve pain of children with generalized dystonia.
Recently, Mohammad et al., outlined a treatment approach for children with MD, including IEMs disorders. In this approach, phenomenology of MD, associated comorbidities (in particular intellectual disability and psychiatric disorders), etiology (and natural history of the disorder), level of functional impairment, and risk of a definite therapy are determined sequentially before making a treatment decision. As in any other MD, the administration of quality of life scales or scales for the specific MD is recommended, in order to have a useful measure of the effect of the treatment. The risk of a given treatment should always be considered before starting, since the adverse effects of the treatment may be worse than the MD itself. Providentially, side effects are transient and reversible for most treatments (149).
Keeping up to date with the emergence of new gene therapies is essential. Gene therapy trials using viral vectors results in improvement of MD and motor function in early clinical studies of AADC deficiency (150,151). Kojima et al., conducted an open-label phase 1/2 study including six patients whom received adeno-associated virus vector harboring DDC via bilateral intraputaminal infusions. Five of these patients presented a severe phenotype while the remaining patient presented a moderate phenotype. At follow-up for up to 2 years, motor function improved markedly in all patients. They were able to stand with support to run or ride a bicycle. The treatment was more effective in the younger patients. Cognitive and verbal functions were also improved in the patient with the moderate phenotype. Clinical signs such as dystonia and oculogyric seizures disappeared or decreased markedly. As an adverse effect, transient chorea was observed (150). Tseng et al., demonstrated that there was an improvement in the microstructural integrity of white matter tracts using Brain diffusion tensor imaging (DTI) 1 year after AADC gene therapy (151).
In the same way, enzyme-replacement therapy for CNL2 have resulted in significant less decline in motor and language function than that in historical controls although with serious adverse events comprising failure of the intraventricular device and device-related infections (152).
For many patients, medical management alone is insufficient. Hence, many of these drug-refractory patients experience significant symptomatic improvement after baclofen intrathecal infusion pump or deep brain stimulation (DBS) following an appropriate selection process. Intrathecal baclofen pump has been indicated in several patients with spasticity or severe dystonia and progressive IEMs, in some cases for palliative purposes, for example: Niemann Pick disease type C (153), Sjögren-Larsson syndrome (154), Metachromatic leukodystrophy (155), and Gaucher disease type 2 (156). Regarding DBS, Supplementary Table 4 describes the role and evidence of DBS in selected IEMs and childhoodonset MD. Twenty-five articles were included with fifty-four patients who underwent DBS surgery. The mean age of surgery was 12.2 ± 4.6 years (range: 1-18 y) and the most frequent targets were internal globus pallidus (19 studies) and subthalamic nucleus (5 studies) . In summary, DBS can be recommended for treatment of dystonia in pantothenate kinase associated neurodegeneration or Lesch-Nyhan syndrome (182). Patients with pantothenate kinase associated neurodegeneration repeatedly respond well to DBS even if there may not be sustained at long-term effect because of ongoing neurodegeneration in the pallidum. While DBS appears effective in Lesch-Nyhan syndrome patients both in reducing dystonia and self-mutilating behavior. No beneficial effect was observed in the following IEMs: X-linked adrenoleukodystrophy, methylmalonic aciduria, mitochondrial disorder and glutaric aciduria type 1.

STEP 6. REEVALUATE TREATABLE IEMs
IEMs presenting with prominent MD can be noticed in Table 1. It is of fundamental importance to reevaluate these treatable IEMs and ensure that they have been completely discarded, both biochemically and genetically. This is a process that must be performed consistently in all patients without diagnosis. Finally, other genetic or acquired causes that could explain the movement disorders will have to be investigated.

FUTURE DIRECTIONS
Further collaborative investigations are needed in which an elaborated description of the MD is made, including the largest cohorts possible. It would also be beneficial if more research was done on the application of intrathecal baclofen pump and deep brain stimulation in IEMs. The future is very promising, both in new diagnostic techniques in particular metabolomics and in new gene therapies. Metabolomics can be described as a complete and simultaneous analysis of small molecules in a biological sample. The aims of this analysis are to improve the diagnosis, understand the factors that contribute to the progression and susceptibility of a certain disease, and monitor the long-term metabolic effects of therapies. The scope and usefulness of this approach in diagnosing IEMs is expected to drastically change in the near future and it is very likely that they will be included in upcoming diagnostic algorithms (183).

CONCLUSION
This review proposes a straightforward six-step algorithm to IEMs presenting with MD in childhood. The major components of this algorithm include the initial suspicion on an IEM, the characterization of the MD phenotype and other neurological and non-neurological signs, biochemical, genetic and imaging testing based on the type of MD. Finally, an update of the treatable IEMs and the therapies available for the MD is made. The application of this algorithm will potentially allow clinicians to avoid blind and poorly efficacious explorations or biological test.

AUTHOR CONTRIBUTIONS
JDO-E wrote and reviewed the final manuscript and provide data for all tables and figures.