Management of neurological symptoms in Lesch-Nyhan disease: a systematic review

: Lesch-Nyhan Disease (LND) is an X-linked recessive genetic disorder arising from hypoxanthine phosphoribosyltransferase 1 gene mutations, leading to a complete deficiency. LND presents a complex neurological profile characterized by generalized dystonia, motor dysfunctions and self-injurious behavior, which management is challenging. We conducted a systematic review of studies assessing the efficacy of pharmacological and non-pharmacological interventions in management of neurological symptoms in LND (PROSPERO registration number

Lesch-Nyhan Disease (LND) is an X-linked recessive genetic disorder arising from hypoxanthine phosphoribosyltransferase 1 gene mutations, leading to a complete deficiency.LND presents a complex neurological profile characterized by generalized dystonia, motor dysfunctions and self-injurious behavior, which management is challenging.We conducted a systematic review of studies assessing the efficacy of pharmacological and non-pharmacological interventions in management of neurological symptoms in LND (PROSPERO registration number:CRD42023446513).
Among 34 reviewed full-text papers; 22 studies were rated as having a high risk of bias.Considerable heterogeneity was found in studies regarding the timing of treatment implementation, adjunctive treatments and outcome assessment.Single-patient studies and clinical trials often showed contradictory results, while therapeutic failures were underreported.
S-Adenosylmethionine and Deep Brain Stimulation were the most studied treatment methods and require further research to address inconsistencies.The evidence from levodopa studies underlines that optimal timing of treatment implementation should be thoroughly investigated.
Standardized study design and reducing publication bias are crucial to overcome current limitations of assessing intervention efficacy in LND.

Introduction
Lesch-Nyhan Disease (LND) is a rare X-linked recessive genetic disorder resulting from mutations in the HPRT1 gene (Nyhan, 1973), leading to a complete enzyme deficiency.The HPRT1 gene encodes the purine salvage pathway enzyme, hypoxanthine-guanine phosphoribosyltransferase (HGPRT), which plays a pivotal role in the conversion of hypoxanthine to inosine monophosphate (IMP) and guanine to guanosine monophosphate (GMP).The classical presentation of LND, characterized by significantly reduced or undetectable HGPRT activity (typically below 1-1.5% of basal activity), manifests with a plethora of symptoms (Jinnah et al., 2006).These include hyperuricemia, motor syndrome, and a variety of neurobehavioral abnormalities, prominently including self-injurious behaviors (SIB) (Jinnah, 2009).Despite decades of research, the neuropathology of LND is still not completely understood.Uric acid accumulation, caused by absent purine salvage, initially speculated to be a contributory factor (Bell et al., 2016), is unlikely to cause neurological symptoms (Jinnah, 2009).Patients with attenuated variants of LND (with HGPRT activity >2%) do not exhibit a full spectrum of neurological and behavioral symptoms, despite hyperuricemia (Fu et al., 2014) (Bell et al., 2016).Moreover, pharmacological interventions targeting uric acid accumulation, such as allopurinol, have been ineffective in alleviating the neurological and neurobehavioral symptoms of LND (Jinnah, 2009).
The limited role of uric acid accumulation and the results of postmortem studies have led to the conceptualization of LND as a developmental disorder caused by impaired energy metabolism in basal ganglia dopaminergic neurons (Bell et al., 2021).Purine salvage pathway deficiency causes prioritization of de novo purine synthesis over energy production, diminishing the developmental capacity of dopaminergic neurons (Bell et al., 2021).Disruption in dopaminergic signaling within the basal ganglia circuits leads to both neurobehavioral and motor symptoms (Baumeister and Frye, 1985) (Göttle et al., 2013) (Visser et al., 2000) (Bell et al., 2016).
The lack of effective therapeutic options leaves most LND patients heavily dependent on wheelchairs, necessitating physical restraint or dental extraction to mitigate SIB (Torres, 2019), and requiring substantial assistance with daily activities (Jinnah et al., 2006).The absence of international treatment guidelines underlines the complexity of managing LND.Current pharmacological approaches primarily rely on allopurinol for uric acid control and medications such as baclofen or benzodiazepines for managing increased muscle tone (Jinnah et al., 2006).However, there is no recommended treatment for motor and neurobehavioral symptoms.
Due to LND rarity, the majority of data focusing on neurological symptoms management originate from case reports and case series, often yielding inconsistent results and subpar methodological quality.This makes evidence-based medical care challenging and necessitates thorough literature searches.In this review, we integrated safety and efficacy profiles for managing neurological symptoms, assessed the methodological quality of studies, and evaluated the evidence for each experimental therapy.We also highlighted promising research areas and suggested improvements for future trials.
J o u r n a l P r e -p r o o f crossover studies and rated according to the specified guidelines (Sterne et al., 2019).Any disagreement in assessment was discussed between rating researchers until consensus was achieved.

Outcome Assessment
Dystonia was assessed with the Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS) in 6 studies, with the Unified Dystonia Rating Scale (UDRS) in 2 studies, and with the Barry-Albright Dystonia Scale (BADS) in a single study.One study used the Abnormal Involuntary Movements Scale (AIMS).Muscle tone was assessed by a single study using the Modified Ashworth Scale.Neurobehavioral symptoms were assessed using the Behavioral Problems Inventory (BPI) in 4 studies and the Adaptive Behavior Scale -Residential and Community Second Edition (ABS-RC-2) in a single study.Additionally, one study used the Child Behavior Checklist (CBCL) and the Adult Behavior Checklist (ABCL).One study each used changes in Positive Mood Rating or Mean Disability Scale (MDS) as the outcome measures.Narrative assessment of treatment or custom quantitative methods were used in the remaining studies.

Treatments primarily targeting motor symptoms
Patients affected by LND exhibit characteristic motor syndrome composed of action dystonia superimposed on baseline hypotonia, sometimes with less prominent choreoathetosis and spasticity (Jinnah et al., 2006).Among potential treatments, only tetrabenazine, and intrathecal baclofen were evaluated to primarily address motor syndrome.

Tetrabenazine
TBZ is a presynaptic dopamine depletor and postsynaptic dopamine receptor blocker that is used to manage hyperkinetic disorders in adults (Jain et al., 2006).One case series of TBZ therapy in children included one LND patient whose motor symptoms, most notably chorea, improved while on 87.5 mg of TBZ (Jain et al., 2006).
J o u r n a l P r e -p r o o f

Intrathecal baclofen
Baclofen is a GABAb receptor agonist, which can be administered orally or intrathecally, and is used to manage spasticity in LND patients (Jinnah et al., 2006).The interplay between dopaminergic and GABAergic systems led to a hypothesis that baclofen may indirectly regulate dopaminergic balance, in turn alleviating neurological symptoms of LND (Pozzi et al., 2014).One study tested intrathecal baclofen in 3 adult LND patients (Pozzi et al., 2014).The treatment was started with an initial dose of 270 µg/day and was titrated until a dose of 550 µg/day.The outcome of treatment was evaluated using the UDRS and qualitative SIB assessment performed by caregivers.Improvements in motor symptoms and SIB were observed (see Table 1.).Adverse effects occurred in only one patient (see Table 1.).One case report described intrathecal baclofen treatment in an adult LND patient (Satow et al., 2021).The initial dose was 24 µg/day and an effect was observed after a week.The UDRS score initially improved, but after a year, it returned to the pretreatment level and remained at it despite increasing the dose to 105 µg/day (see Table 1).

Treatment methods primarily targeting neurobehavioral symptoms
The neurobehavioral component of LND is characterized by the presence of SIB, which is defined as non-accidental behavior resulting in demonstrable, self-inflicted physical injury without the intent of suicide or sexual arousal (Huisman et al., 2018).In LND patients, SIB consists of a distressing set of behaviors including self-biting, head-banging, eye-poking, and wheelchairtipping (Robey et al., 2007).SIB in LND can result in severe tissue damage and disfigurement of the face and tongue (Torres, 2019).Due to the symptom severity and the lack of effective treatment, SIB in LND is managed with restraints and dental interventions.These measures only preclude SIB instead of eliminating them (Olson and Houlihan, 2000).Furthermore, they restrict patients' activity (Olson and Houlihan, 2000) and negatively affect their' psychosocial well-being (Isola et al., 2022).The failure rate of devices to prevent self-biting is high, and patients may develop new self-injury methods after dental extractions (Torres, 2019).Hence, restraints and dental extraction are not the preferred methods for managing SIB (Olson and Houlihan, 2000), but their use may be necessary and should not be delayed in severe cases (Goodman et al., 2014).The neurobehavioral component can also consist of aggressiveness, anxious-depressive symptoms, and interpersonal difficulties (Schretlen et al., 2007), although the reviewed studies rarely evaluated these aspects.

Levodopa/Carbidopa
LND is associated with a deficit of dopaminergic neurons and dysfunction of basal ganglia motor circuits (Visser et al., 2000) (Bell et al., 2021) (See Fig 3).Furthermore, levodopa efficacy was demonstrated in treating dystonias caused by decreased dopamine levels such as in GTP cyclohydrolase deficiency.It was speculated that restoring dopamine balance could benefit LND patients (Visser et al., 2011).
One report on levodopa/carbidopa treatment of an infant diagnosed with LND, who had not developed SIB yet, was included (Serrano et al., 2008).Levodopa/carbidopa was administered in a proportion of 1/0.25 in a dose up to 3 mg/kg/day, which improved motor symptoms.By the last follow-up at the age of 3, SIB had not developed.The treatment was continued with the dose titrated to 6 mg/kg/day.A prospective study evaluated the efficacy of levodopa/carbidopa in 5 patients whose ages ranged from 3 to 27 years (Visser et al., 2011).The study was terminated preemptively in the titration phase due to the exacerbation of motor symptoms in 4 out of 5 patients (see Table 1).The average daily dose of levodopa/carbidopa was 200/50 mg.

J o u r n a l P r e -p r o o f
One study evaluated levodopa/carbidopa in three infants aged 11 to 13 months (Visser et al., 2024).Levodopa/carbidopa was administered in a ratio of 1:0.25 at dose escalated up to 6 mg/kg/day, divided into three doses.Two of the three patients did not develop self-injurious behavior (SIB) by the ages 15.5 and 14 years, respectively.The third patient developed SIB at the age of one and a half years while on a dose of 2.5 mg/kg/day of levodopa/carbidopa.All three patients developed motor syndrome.

S-adenosylomethionine
SAM is a purine derivative which is a universal methyl group donor, thus regulating replication, transcription, and translation (Loenen, 2006).Following the donation of its methyl group, SAM can be further metabolized to adenosine, potentially replenishing the purine pool in LND (Chen et al., 2014) (See Fig. 3).In addition to its involvement in transsulfuration and polyamine synthesis (Loenen, 2006), SAM impacts monoamine neurotransmitter metabolism (Bottiglieri et al., 1994).After oral or intravenous administration, SAM levels increase in the cerebrospinal fluid (CSF) (Bottiglieri et al., 1990).However, the exact transport mechanism of SAM remains elusive (Chishty et al., 2002).Furthermore, a preclinical study identified SAM as a potential rescuing compound for LND (Ruillier et al., 2020).Initially, SAM was administered to adult LND patient orally at a dose of 400 mg twice/day to reduce elevated levels of transaminases (Glick, 2006).Due to the unexpected improvement in the patient's mood and SIB, the dose was increased to 800 mg twice/day which caused SIB cessation within 18 months.Irritability was assessed using the Percent Positive Mood Rating and after a year an improvement from 36% to 58% was achieved.The 3-year follow-up reported further improvement to 96% on the Percent Positive Mood Rating.Two additional case reports described patients who were treated with SAM.Initially, SIB of the first patient was controlled using preventive measures, including a mouthguard, but the worsening of the symptoms led to the introduction of SAM (Narai et al., 2021).The dose was titrated from 400 mg/day to 1000 mg/day (40 mg/kg/day) which resulted in alleviation of SIB.By the last follow-up at the age of 16, the improvement persisted and the dose of SAM was increased to 1200 mg/day.The second patient was 26 months old when treatment started with oral SAM at a dose of 20 mg/kg/day (Momosaki et al., 2020).Gradual improvements in motor symptoms were observed.SIB alleviated and resolved completely after introducing risperidone at a dose of 0.05 mg/kg/day as a supplementary treatment.Another case report describes a patient treated with SAM orally at a dose of 100 mg twice/day (17 mg/kg/day) (Lauber et al., 2017).The study consisted of an observational period, a sequence of placebo-verum periods, and finally a continuous therapy with SAM.SIB incidents were recorded and counted.In verum periods the mean number of SIB incidents per day was 4.57 and in placebo periods it was equal to 8.62 per day.An open-label prospective trial on SAM treatment enrolled 5 children, who received doses ranging from 22 to 38 mg/kg/day (Chen et al., 2014).The improvement was estimated by the caregivers (see Table 1).Another open-label dose escalation trial enrolled 14 patients, of whom 13 were adults (Dolcetta et al., 2013).Three patients in this study were diagnosed with attenuated variants of LND, without SIB, but exhibiting anxiety.The initial SAM dose was 400 mg/day orally and the planned maximum dose was 800 mg twice/day.The treatment was withdrawn in 10 out of 14 patients in the titration phase due to the occurrence of adverse effects (see Table 1).The remaining 4 patients continued treatment till the maximum dose.Among responsive patients, 3 presented with classical LND, and 1 patient with an attenuated variant whose neurobehavioral component was limited to anxiety.For all responsive patients, an improvement in mood was reported and the frequency of SIB incidents decreased in patients with classical LND.
J o u r n a l P r e -p r o o f

5-Aminoimidazole-4-carboxamide ribonucleotide
AICAR is a metabolite of the de novo purine synthesis pathway, which has been shown to normalize levels of adenine and guanine in HGPRT deficient cells, compared to those found in normal cells (Page, 1989).Hence, it was hypothesized that AICAR may compensate for HGPRT deficiency in patients with LND (Page et al., 1995) (See Fig. 3).One study tested AICAR efficacy in 2 LND patients (Page et al., 1995).The first patient was 16 years old and received 30 mg/kg/day of AICAR orally for 4 days, followed by an increase in dosage to 100 mg/kg/day for the subsequent 4 days.The second patient was 29 years old and received daily infusions of AICAR until a plasma level of 50 µM was achieved.AICAR was not detected in the CSF.No improvement in neurological symptoms was observed in either patient.

Gabapentin
Gabapentin was initially synthesized to manage spasticity but is currently used as an adjunctive anticonvulsant and co-analgesic due to its antinociceptive, anxiolytic, and neuroprotective properties (Taylor et al., 1998).Gabapentin modulates the activity of glutamic acid decarboxylase and thus increases the production rate of gamma-aminobutyric acid (GABA) in the brain (Taylor et al., 1998), and it is speculated that it also increases serotonin levels in the central nervous system (McManaman and Tam, 1999).This could be beneficial for LND patients, who have altered levels of these neurotransmitters (Saito and Takashima, 2000).Additional proposed mechanisms of gabapentin action in LND are reducing endogenous convulsants (kynurenine metabolites) and increasing endogenous anticonvulsants (serotonin and GABA) (Gedye, 1992).Other hypotheses of gabapentin mechanism of action include interactions with voltage-sensitive calcium channels and influencing the release of monoamine neurotransmitters (Taylor et al., 1998).Furthermore, gabapentin is beneficial in the treatment of some motor disorders such as Parkinson's disease (Olson et al., 1997) or amyotrophic lateral sclerosis (Mazzini et al., 1998), can be used to manage pain (Kukkar et al., 2013) and spasticity (Priebe et al., 1997).Besides, gabapentin is used as a mood stabilizer and antidepressant in LND patients (Jinnah et al., 2006).A case report presented a patient who was initially treated with 400 mg/day gabapentin which resulted in noticeable SIB reduction (McManaman and Tam, 1999).Subsequently, the gabapentin dose was increased to 800 mg/day, and after 3 weeks of treatment, parents reported complete cessation of SIB.Due to the necessity of treating a bowel obstruction, gabapentin was withdrawn and SIB recurred.After reinitiating the treatment with gabapentin, SIB immediately improved

Carbamazepine
Carbamazepine is an anticonvulsant that is also used as a mood stabilizer in neuropsychiatric disorders (Smith, 2005).Besides, it has demonstrated effectiveness in managing SIB exhibited by patients diagnosed with intellectual disabilities (Barrett et al., 1988) or acquired sensory neuropathy (Roach et al., 1996).Based on the resemblance of those symptoms to those in LND, (Roach et al. 1996).tested the efficacy and safety of carbamazepine in managing SIB in LND patients.The baseline SIB severity of four patients, including one adult was assessed during 5 days of inpatient observation.Treatment was initiated with 10 mg/kg/day, which was steadily increased to achieve serum levels of carbamazepine between 6 to 10 µg/mL, while patients remained 5 days of inpatient observation to assess treatment efficacy.Behavior was also assessed by parents and school personnel.SIB of 3 patients improved.To confirm the link between treatment and effect, carbamazepine was withdrawn in 2 patients, and SIB recurred till treatment was J o u r n a l P r e -p r o o f restarted.One patient was transiently responsive to the treatment only after reaching a dose of 26.6 mg/kg/day and a serum level of 13 µg/mL, which led to the treatment withdrawal.

Selective Serotonin Reuptake Inhibitors
SSRIs are primarily used in treatment of depression and act by modulating the connectivity of neural networks critical for action selection and emotional processing (Wagner et al., 2017) (Kim et al., 2020) (Roberts et al., 2020).Kirkpatrick-Sanchez et al., (1998) tested 2 different SSRIs -paroxetine and sertraline, simultaneously with a "serotonergic diet" to manage SIB in LND patient -The serotonergic diet consisted of supplementing daily meals with serotonin-containing foods 6 times a day, based on a report of successful treatment of SIB caused by 21-chromosome trisomy using supplementation (Gedye, 1990).Paroxetine was started at 5 mg/day and titrated up to 30 mg/day, while sertraline was started at 12.5 mg/day and titrated up to 200 mg/day.The serotonergic diet alone did not have any effect on SIB, but both paroxetine and sertraline reduced -non-restrained SIB attempts.The most effective dosage was 20 mg/day for paroxetine and 75 mg/day for sertraline.

Risperidone
Risperidone is an atypical antipsychotic primarily indicated in schizophrenia, however, it is used in the short-term management of behavioral disturbances in autistic disorder (Pandina et al., 2007) (McCracken et al., 2002), Tourette syndrome (Dion et al., 2002) and Cornelia De Lange syndrome (Kline et al., 2018).Since risperidone is a serotonin-dopamine antagonist it was hypothesized that it might have potential beneficial effect on neurobehavioral symptoms observed in LND (Allen and Rice, 1996).Indeed a beneficial effect of risperidone was observed on SIB of an adult LND patient (Allen and Rice, 1996).Risperidone was initiated at a dose of 2 mg twice/day.Compared to previously collected baseline data improvement in monthly instances of SIB incidents was observed.Treatment with both SAM and risperidone also led to cessation of SIB in one patient (Momosaki et al., 2020).

Ecopicam
Ecopicam is a selective D1 antagonist that attenuated SIB in a 6-hydroxydopamine animal model of dopamine depletion (Criswell et al., 1992) and effectively reduced tic frequency in Tourette syndrome (Gilbert et al., 2023).
An open-label dose escalation trial with 5 patients was conducted to assess ecopipam safety in LND patients (Khasnavis et al., 2016a).The initial dose of 12.5 mg/day was increased throughout the study.The planned final dose of 200 mg/day was reached by one patient according to the study schedule.Two patients had their doses increased more steadily, with a final dose of 100 mg/day, and the remaining two patients had their doses lowered due to adverse effects (see Table 1).A reduction in the number of out-of-restraint SIB incidents per minute was observed, but the study was underpowered to assess efficacy.A randomized double-blind study of ecopicam, which was designed to assess efficacy, omitted the titration phase and was terminated due to the occurrence of numerous adverse effects (see Table 1) (Khasnavis et al., 2016b).The patients had the option to continue the treatment in an open-label phase and the only patient who continued had a major reduction in the frequency of SIB incidents.
J o u r n a l P r e -p r o o f

Botulinum Toxin Injections
Botulinum toxin exerts a temporary inhibitory effect on the release of presynaptic acetylcholine, leading to muscle weakness (Dabrowski et al., 2005).There are several botulinum toxin serotypes (A-G), which differ in potency, targeted protein, and additional mechanisms of action (Aoki and Guyer, 2001).Botulinum-A serotype was used in all included studies to weaken the muscles implicated in SIB (See Fig. 3) and subsequently reduce SIB frequency and severity (Dabrowski et al., 2005) (Garcia-Romero et al., 2022) (Gilbert et al., 2021) (Gutierrez et al., 2008).The injection targets are masticatory muscles, which are involved in self-biting, and the muscles involved in kicking or beating (Garcia-Romero et al., 2022).Two LND children were treated with botulinum toxin injections.One patient was injected into the bilateral masseters at 2 sites at a dose of 40 units per muscle group (Dabrowski et al., 2005).The parents reported a decrease in selfbiting and aggressiveness.The effects of each injection lasted for 10 weeks, after which symptoms returned to baseline level and injections had to be repeated.The injections were repeated every 12 weeks, which led to a total number of 3 visits during the study timeframe.The patient returned to school and sores on his hands, mouth, and tongue, which were caused by self-biting, healed.The second patient received botulinum injections into masseter muscles at a dose of 20 units per side (Gilbert et al., 2021).The 6 and 16-week follow-ups reported SIB alleviation.Five months after injections, SIB recurred.Therefore injections were repeated 3 more times every 5 months at a dose of 25 units per side with symptom relief after each injection.An improvement in speech, evaluated with Photo Articulation Test Third Edition, was also observed.The botulinum toxin therapy was tested in an adult LND patient as well (Gutierrez et al., 2008).He received botulinum injections into zygomatics, lip orbicularis, and mentalis muscles.The zygomatics were injected bilaterally with 12.5 units each, orbicularis, with 6 injection sites, 2.5 units each.Levator labii inferioris was injected in 3 sites, 5 units each.Injections were repeated twice with a 3-month interval.The frequency of self-biting decreased, which allowed the mouth injuries caused by them to heal.One study retrospectively analyzed the safety and efficacy of botulinum toxin therapy in 6 patients (Garcia-Romero et al., 2022), including one adult.Apart from botulinum injections, all patients received additional pharmacological treatment to control SIB (see Table 1).Periods during which patients received the injections varied from 1.5 to 7.1 years.Botulinum toxin was injected bilaterally into the masseter, temporalis, and biceps brachii.Two patients were additionally injected in the quadriceps muscle to reduce kicking.The initial doses ranged from 8 to 15.8 units/kg and maximum doses ranged from 12 to 30.6 units/kg.The patients received 119 injections in total, and the mean number of injections received by each patient was 20.Only 3 injections were associated with adverse effects.These were 2 instances of temporal difficulty with swallowing and one instance of general weakness which disappeared less than a week after injection.Injections were considered completely effective if self-biting was reported to disappear and as partially effective if the number of self-biting events decreased (see Table 1).The frequency of self-biting was reported to reduce, although restraints remained in use most of the time.On the contrary, other forms of self-injury which included the use of arms or legs were unaffected by the treatment.The mean time of single injection effectiveness was 69.1 days.Two patients, who benefited from the treatment had to undergo tooth extraction, as tooth eruption coincided with the reappearance of self-biting, despite increasing the botulinum toxin dose.

3.5
Treatments targeting both motor and neurobehavioral symptoms.

Deep Brain Stimulation.
J o u r n a l P r e -p r o o f Deep Brain Stimulation (DBS) of GPi has long been a form of treatment of dystonia (Tambirajoo et al., 2021) although the exact mechanism behind beneficial outcomes remains unknown.It is known that adenosine release caused by stimulation is essential for therapeutic effect (Bekar et al., 2008), which may be of importance for LND treatment.In the case of LND, single or double leads are applied bilaterally into GPi (Deng et al., 2023).Leads can be placed in the anteroventral part of GPi (limbic stimulation) and/or in the posteroventral (motor stimulation) part.DBS was first used in an LND patient to primarily control dystonic movements (Taira et al., 2003).Single leads were applied bilaterally into GPi.The patient's BFMDRS score showed 33% improvement in the movement section and 50% in the disability section.SIB completely disappeared.In a subsequent study a 16-year-old patient (Cif et al., 2007), had double leads implanted bilaterally into anteroventral and posteroventral parts of GPi.To assess the specific influence of each lead on behavioral and motor symptoms, authors conducted tests of selective stimulation.Disabling limbic (anteroventral) stimulation caused the BPI score to worsen, with a small improvement compared to baseline.Lack of limbic stimulation did not affect the BFMDRS score on the motor section.Lack of motor (posteroventral) stimulation caused the BFMDRS score to worsen and slightly affected the BPI score.A 40% improvement in BFMDRS was found in the movement section with all leads active.SIB disappeared within several days after the double stimulation started.In a 10-year-old boy intrathecal baclofen treatment had to be terminated due to infections (Deon et al., 2012), and was subsequently treated with DBS stimulation.His dystonia improved as measured with BADS.Furthermore, SIB disappeared and muscle tone decreased.The improvements persisted and no complications were reported 2 and a half years after the implantation procedure.A 15-year-old, who despite dental extraction and application of restraints, was able to elicit self-injury was treated with DBS (Abel et al., 2014).He got leads implanted bilaterally into GPi, and subsequently, the frequency of SIB incidents decreased to the point of enabling complete restraint removal.Furthermore, the patient's dystonia improved.Several months after surgery the patient fell, fracturing the right lead which coincided with aggravation of dystonia and SIB on the left side of the body.Both leads were replaced, with the right one being placed 2 mm deeper and more posteriorly compared to the original location and the left being implanted in the same place.His dystonia improved but was noticeably worse on the left side of the body.SIB persisted, but restraints were unrequired.In an adult with severe LND single leads were implanted bilaterally (Piedimonte et al., 2015).The results of the surgery were assessed in 12 separate measurements taken up to 60 months after treatment.After intervention, both BFMDRS and.MDS scores improved (see Table 1).Four months after stimulation started, SIB has disappeared completely.Disabling the stimulation caused the therapeutic effect to diminish till stimulation was reinstated.For 4 male children initial improvement was observed in both dystonia and behavioral symptoms (Tambirajoo et al., 2021).The first patient had double leads placed bilaterally into anteromedial and posteroventral GPi.However, an infection of the internal pulse generator (IPG) site forced the removal of the entire system which caused regression of the therapeutic effect.In the 8-year follow-up, his BFMDRS improved by 3.8% in movement and 6.9% in disability sections.The BPI score improves by 68.1% in frequency and 78.1% in severity.The second patient already had single bilateral leads implanted in a different center, but due to the worsening of his symptoms, IPG and external cables were replaced.A few years later his symptoms worsened again and the whole system was replaced with bilateral double leads.Initial improvements suddenly disappeared which was interpreted as a fault of external cables, and replacing them restored the beneficial effect.In the 12-year follow-up, his BFMDRS improved by 1.3% on movement and 4% on disability.BPI improved by 53% in frequency and 50% in severity.

J o u r n a l P r e -p r o o f
The third patient had bilateral double leads implantation into GPi.Four months after surgery, an infection in the IPG site required system removal and antibiotic therapy.Three years later the new system was implanted.Three years post-surgery his BFMDRS improved by 6.7% on movement and 8.3% on disability.The fourth patient underwent implantation of bilateral single lead into posteroventral GPi.The rationale for implanting single lead was due to the risk of infections.The patient's baseline BFMDRS score on the motor section was 93.5 and 26 on the disability section.The 1.8-year followup showed an improvement of 1.6% in the movement and no change in the disability section.Recurrent infections led to the removal of the entire system.Another case series of 2 pediatric patients who underwent DBS implantation into GPi reported improvements in SIB and motor symptoms (Pralong et al., 2005).A retrospective assessment of 14 patients showed numerous complications of DBS treatment (Visser et al., 2021).Data was collected by surveying patients and their families with a questionnaire for on average 5 and a half years after the intervention.All but one participant reported various adverse effects of the treatment.Furthermore, 6 patients reported infections of either scalp or leads and 5 patients reported hardware malfunctions.Only 2 patients did not require any additional operations.About half of both patients and caregivers were positive about the outcome of treatment.

Bone Marrow Transplantation
Bone marrow transplantation was speculated to alleviate symptoms of LND (Endres et al., 1991).However, in one LND patient receiving allogeneic bone marrow transplantation (Endres et al., 1991) no improvements in neurological symptoms were observed, while the patient died 10 days post-transplantation due to post-procedural complications.

Allogeneic Umbilical Cord Blood Transplantation
A two-year-old patient received allogeneic umbilical cord blood transplantation.This patient had not yet developed SIB (Kállay et al., 2012), but following the transplantation, motor symptoms, and irritability worsened in addition to several adverse effects.The authors reported small mental improvement but no significant motor improvement after transplantation.The threeyear follow-up reported that SIB had not developed, and the patient was able to walk with support and speak in short sentences.

Allopurinol
Allopurinol is widely used in LND patients to manage hyperuricemia by inhibiting xanthine oxidase (Jinnah, 2009).Although some authors suggest it crosses the blood-brain barrier easily (Karageorgos et al., 2006), its target is nearly absent in the brain (Saksela et al., 1998), and allopurinol measurements in patient CSF indicate otherwise (Murrell & Rapeport, 1986).A study with a low risk of bias and 18 participants reported no effect of oral allopurinol on neurological symptoms (Torres et al., 2007).

Discussion
The lack of effective treatment for motor syndrome and SIB observed in LND makes it a heavy burden for both patients and their caregivers.A comprehensive review of available treatment options targeting neurological symptoms in LND is lacking, apart from one paper on DBS (Deng et al., 2023) and an expert recommendation (Torres, 2019).

J o u r n a l P r e -p r o o f
This review identified 2 interventions that were studied the most extensively -SAM and DBS (see Fig. 2).The former is reported as an effective treatment of SIB in retrospective studies (see Fig. 2), albeit one open-label study identified responders and non-responders groups, with no clear explanations responsible for the split (Dolcetta et al., 2013).The same study reported anxiety or sleep disturbances as common side effects.Out of the six included studies, only two were assessed as low risk of bias (see Supplementary Table 2A).One was a single-patient prospective study with successful treatment outcomes (Lauber et al., 2017), while the other found that most patients were non-responders and experienced adverse effects (Dolcetta et al., 2013).Although single-patient reports suggest that SAM treatment might improve symptoms, the majority of reviewed patients did not benefit (see Table 1).Further research is necessary before SAM can be recommended for LND treatment.Investigating the mechanisms of SAM's efficacy and tolerability in LND patients is essential to resolve these conflicting results.Due to the short half-life of SAM (Giulidori et al., 1984), investigators should monitor SAM plasma levels during treatment, and consider the compound instability when exposed to oxygen (Young and Shalchi, 2005).Furthermore, intravenous administration of SAM in LND patients should be considered.All reports on DBS stated its effectiveness in controlling SIB, although the improvement in dystonia measured by BFMDRS varied from 55% (Piedimonte et al., 2015) to 1.3% (Tambirajoo et al., 2021).Three out of eight studies had a high risk of bias, One of the low-risk-of-bias studies reported multiple adverse effects of the intervention (Visser et al., 2021).The main disadvantages of this intervention are a high risk of infection and hardware malfunction, which required additional procedures in 86% of patients in one study (Visser et al., 2021).The occurrence rate of postprocedural complications is higher in LND patients compared to other dystonic patients (Deng et al., 2023); the reason for this is still unclear and needs to be clarified to avoid complications in the future.The patientcentered survey (Visser et al., 2021) illustrates the essence of publication bias in LND when encouraging case reports are confronted with the long-term patient and primary caregivers' experiences.Longer follow-ups in DBS studies are necessary as the initial beneficial effect might be unsustainable due to postprocedural complications.Taking into account the lack of effective treatment for neurological symptoms of LND and their severity, DBS is an option requiring further research due to its potential efficacy.The variable effect on dystonia and the increased adverse effect ratio in LND compared to the rest of the pediatric population must be investigated, longterm improvement in patients' quality of life must justify undertaking surgical treatment.
The third most studied treatment option was botulinum toxin injections.All studies reported botulinum injections effective in controlling self-biting, although only two out of four studies were rated as having a low risk of bias.Other forms of SIB did not improve with this treatment.One study that did not meet inclusion criteria reported the botulinum treatment ineffective (Goodman et al., 2014), so publication bias must be considered.The rate of reported side effects was low but should be always considered as the difficulties with breathing and dysphagia are known adverse effects of botulinum toxin (Gilbert et al., 2021).These can be especially dangerous for LND patients as dysphagia and respiratory events are known causes of sudden death in this population (Neychev and Jinnah, 2006).Botulinum may be useful as a supplementary treatment to manage self-biting, but it does not affect other forms of SIB.Although botulinum injections were only used to manage SIB, this therapy is also an effective treatment option for dystonia (Dressler et al., 2021) and spasticity (Simpson et al., 2016).Therefore, botulinum toxin treatment could be beneficial to control motor symptoms of LND.
Levodopa/carbidopa was reported beneficial in a single case report (Serrano et al., 2008) and case series (Visser et al., 2024).Both studies stated that levodopa/carbidopa prevented the occurrence of SIB if the treatment is started in infancy, with the latter study having sufficient J o u r n a l P r e -p r o o f follow-up by the age of 15.5 and 14 years (Visser et al., 2024).Further trials are needed to elucidate the time window for levodopa/carbidopa introduction and confirm its efficacy in preventing SIB.However, levodopa/carbidopa is associated with exacerbation of motor symptoms and poor reliability in addressing symptoms of adult LND patients, as reviewed elsewhere (Jinnah et al., 2006) and shown in a prospective study (Visser et al., 2011), therefore it should be avoided in this population (Torres, 2019).Other dopamine receptors agonists, ropinirole and pramipexole were found ineffective (Visser et al., 2011).
Two reports of intrathecal baclofen suggest that it is beneficial in managing spasticity, dystonia (Pozzi et al., 2014) (Satow et al., 2021) and potentially alleviating SIB (Pozzi et al., 2014).In one report having a high risk of bias, the patient regressed to the pretreatment BDFMRS score during treatment (Satow et al., 2021).Thus, only one study reports persistent improvement in LND patients treated with intrathecal baclofen (Pozzi et al., 2014).This treatment is associated with a risk of infection which is the reason for frequent pump explantation (Dickey et al., 2013).This problem could be potentially solved through oral administration of baclofen, but due to the inferior bioavailability, it might be inefficient in managing the symptoms (see Table 1).Moreover, intrathecal baclofen requires high doses in the range of 270-550 µg/kg (Pozzi et al., 2014) or the therapeutic effect may not be maintained (Satow et al., 2021).Moreover, high doses may increase the rate and severity of adverse effects so that dose escalation trial of intrathecal baclofen should be conducted carefully.Although ecopicam caused numerous adverse effects in one study (Khasnavis et al., 2016b), it is potentially beneficial for LND patients (Khasnavis et al., 2016a) (Khasnavis et al., 2016b) and a trial with adjusted doses should be undertaken.
Apart from allopurinol, two other treatment methods were exclusively reported to be ineffective (see Fig. 2).AICAR did not affect the symptoms of two patients (Page et al., 1995), which could be caused by its limited ability to cross the blood-brain barrier (Marangos et al., 1990).A study of Aminoimidazole Carboxamide Riboside was registered as NCT00004314 in ClinicalTrials.gov,but no results were published.Allogeneic bone marrow transplantation led to a patient's death due to postprocedural complications (Endres et al., 1991).Furthermore, another study reported no beneficial effect of bone marrow transplantation in an LND patient (Nyhan et al., 1986).Preclinical research on rodents suggests a lack of impact of bone marrow transplantation on the HGPRT deficient phenotype (Wojcik et al., 1999), and this treatment method is unsupported by available data (Torres, 2019).
Other treatment methods were presented as isolated reports (see Fig. 2).Gabapentin, risperidone, and carbamazepine were reported to be effective in single studies with a high risk of bias (see Fig. 2) (Allen and Rice, 1996) (Roach et al., 1996) (McManaman andTam, 1999).However, their frequent use in LND patients was observed (see Table 1).Gabapentin was mentioned as a prior treatment in 7 (21.2%),risperidone in 7 (21.2%),and carbamazepine in 5 (15.2%) of the included studies (see Table 1).These drugs were rarely considered effective with only a minor effect on behavior (Torres, 2019), however, they were used as a supplementary treatment (Torres, 2019) (Anderson andErnst, 1994).Their risk-benefit ratio should be reevaluated.TBZ was reported to be effective in treating motor symptoms in one patient (Jain et al., 2006), but based on expert opinion the majority of patients were intolerant to TBZ (Torres, 2019).Sertraline and paroxetine were reported to reduce SIB in one report with a high risk of bias (see Fig. 2), therefore their efficacy cannot be concluded.
The future of LND treatment relies on our better understanding of the disease mechanisms (see Fig. 3).One study developed a screening assay with iPSCs derived from LND patients and identified 6 potential drug candidates (Ruillier et al., 2020).All identified compounds have at least J o u r n a l P r e -p r o o f one adenosine moiety in their structure, and include SAM which has been already tested in LND patients (see Fig. 2).Among other hit compounds, nicotinamide adenine dinucleotide (NAD+) and nicotinamide adenine dinucleotide phosphate (NADP) could be administered as either in their native form or as their precursors such as nicotinamide riboside (NR), which can increase their bioavailability in the brain (Lautrup et al., 2019) (Vreones et al., 2023).The experimental models should be further optimized with the use of neural organoids cultured from iPSCs (Vandana et al., 2023), which would enable both a high throughput drug screening and investigation of disease mechanisms.
At the clinical level, studies must employ more rigorous design choices due to the rarity of the disease and limited data.Investigators should take into account that LND is usually diagnosed around 2-3 years of age when SIB emerges (Schretlen et al., 2007).Most of the investigated patients were older (see Fig. 2), which generates uncertainty regarding the early implementation of treatment, as LND is a neurodevelopmental disease, the occurrence of the symptoms could be halted when treatment starts in the first years of life.Levodopa/carbidopa studies strengthen the importance of early implementation of the treatment (see Fig 2).Furthermore, SIB can emerge as late as the age of 10 years (Robey et al., 2003) which must be taken into account during the interpretation of study results.Crossover double-blind design seems to be the most applicable due to the small sample size and has been already used in two studies (Khasnavis et al., 2016b) (Lauber et al., 2017).The design recommendations for N is 1 studies in rare diseases were extensively reviewed elsewhere (Müller et al., 2021), and should be followed due to the rarity of LND and the high risk of bias of many published studies (see Fig 2).We recommend pre-registering all planned trials in LND to increase transparency; underrepresentation of published therapeutic failures can lead to suboptimal clinical choices (Goodman et al., 2014).Considering publication bias and insufficient follow-ups, it would be beneficial to undertake patient-centered surveys to assess longterm treatment outcomes as demonstrated in DBS (Visser et al., 2021).In case of SIB, withdrawing a drug that seems to reduce severe symptoms seems unethical (Rizvi and Nock, 2008), which may end up in a therapy with no clear rationale (Baumeister and Sevin, 1990).An alternative solution is a slow reduction of drug intake to a minimal effective dose instead of withdrawing it completely (Gualtieri and Schroeder, 1990).One study identified a strong placebo effect on SIB from being enrolled in a clinical study, which strongly questions the reliability of uncontrolled studies (Khasnavis et al., 2016b).The other identified problem is the collection of data, which is frequently qualitative (see Table 1).There are no validated scales specifically for LND, but BFMDRS and BPI are most commonly used for dystonia and neurobehavioral symptoms, respectively, and therefore could be the best option for increasing result comparability.One different approach is a quantitative assessment of SIB events in given periods.This method of quantification permits the use of statistics in evaluating treatment effectiveness but raises questions about accurately defining and recording SIB.Validation of scales for LND symptoms would be crucial in reducing bias stemming from symptom severity and frequency assessment.

Limitations
The rarity of LND and attenuated variants contribute to a limited number of studies, often uncontrolled and usually with small sample sizes.Inter-study comparison is a serious significant challenge due to substantial variations in scales and protocols utilized for outcome assessment.Multiple studies only provide narrative descriptions of treatment effects which increases the risk of bias.An additional risk of bias is found in studies that gather treatment efficacy data only from primary caregiver reports.In addition, the varied multi-drug regimens complicate the J o u r n a l P r e -p r o o f interpretation.This aspect hinders the comparison of treatment efficacy between patients and hampers the establishment of causal relationships between specific drugs and their therapeutic effects.Due to the rare occurrence of LND, most studies found were case reports, which are prone to publication bias.Only three single-patient reports assessed the tested treatment as ineffective (see Fig. 2).Seven out of 14 studies (50%) with more than one patient reported the tested treatment as ineffective or with mixed results (see Fig. 2).Despite rare reports of failed interventions, most studies indicate that patients received ineffective or insufficient treatment prior to the intervention (see Table 1).Only two included studies (Khasnavis et al., 2016a) (Khasnavis et al., 2016b) were registered on ClinicalTrials.gov,both reporting mixed effects of the intervention.
We cannot provide strong clinical recommendations based on the existing literature.However, we identified common methodological flaws and proposed areas for improvement in future trials.Studies published before 1990 were excluded from data synthesis, though some preliminary results from earlier studies were discussed.This review did not cover interventions such as restraints or devices for preventing oral self-injuries, as they preclude SIB rather than eliminating it.

Conclusions
There is a need for standardization in future LND trials, either through the universal adaptation of one of the scales or by providing quantitative data such as SIB frequency to facilitate inter-study comparability.The high heterogeneity hinders our understanding of the link between treatment and effect.The differences in timing of treatment, supplementary treatments, and methods of assessment lead to conflicting outcomes.Trials design must be rigorous to minimize the risk of bias.Therapeutic failures were underreported; studies interpretations require caution.
Future studies should specify all drugs and either withdraw ineffective agents or attempt to assess the most effective drug scheme.This review identified SAM and DBS as the most studied treatment methods.Further prospective studies of these treatment options should be conducted to resolve conflicting results of single-patient and multi-patient trials.Botulinum injections appear as a relatively safe and efficient supplementary treatment for self-biting, although no prospective study on this treatment has been conducted to date and only two studies were assessed as low risk of bias, further trials are required.Furthermore, even though severe adverse effects were unreported in the included studies, they must be considered.Introducing levodopa/carbidopa in infancy has been shown to prevent SIB and requires further investigation.Adenosyl-moiety containing compounds, identified by preclinical research, offer promising prospects for the future of LND treatment.

Tables
J o u r n a l P r e -p r o o f

Figure 1 .
Figure 1.PRISMA flow diagram of study selection process.

Figure 2 .
Figure 2. Reported clinical outcomes for included methods of management of neurological symptoms in Lesch-Nyhan Disease.Each shape represents individual study.The studies having a high risk of bias are represented by a triangular shape, studies having low risk of bias are represented by a circular shape.Abbreviations: AICAR, 5-Aminoimidazole-4-carboxamide ribonucleotide; DBS, Deep Brain Stimulation; SAM, S-adenosylmethionine; SSRIs, Selective Serotonin Reuptake Inhibitors.

Table 1 .
Summary of studies characteristics, main clinical outcomes, and adverse effects.