Autism spectrum disorder (ASD) is a set of neurodevelopmental disorders that are characterized by deficits in social interaction and communication, restricted interests, and repetitive behaviors. ASD is often accompanied by psychiatric disorders such as anxiety or attention-deficit hyperactivity disorder [1,2], both of which contribute to sleep disturbances [3] like difficulty falling asleep, frequent nocturnal awakenings, short duration of sleep, early morning awakenings, and daytime sleepiness [4]. Sleep is particularly important in children, and inadequate or disrupted sleep can significantly impact neurodevelopment [5]. Poor sleep is linked to increased aggressiveness, irritability, tantrums, and hyperactivity in patients with ASD [6]. Studies have pointed to a potential bidirectional relationship between sleep disturbances and the severity of ASD [3,4,7], further stressing the importance of adequately managing sleep in these patients.

Sleep disturbances in children with ASD can be initially managed with sleep hygiene and behavioral therapy. If these therapies fail, pharmacological intervention can be considered. Off-label use of antipsychotics, antidepressants, and anticonvulsants constituted the only regimens with approved drugs that were used until recently [8]. Melatonin is the main hormone produced by the pineal gland at night and is involved in the sleep-wake cycle [9]. Exogenous melatonin has proven effective in treating sleep disturbances (mainly sleep onset) in patients with ASD [8], but given its short half-life (45 to 65 minutes) [9], high doses of the immediate-release formulation are needed to achieve high concentrations throughout the night [8]. Immediate-release melatonin is more suitable for inducing sleep and phase shift, whereas prolonged-release melatonin is also effective in maintaining sleep [9]. The regulatory status of melatonin varies between countries. While sold as a dietary supplement in the USA and Canada, melatonin requires a prescription in Europe, where only pediatric prolonged-release melatonin (PedPRM) is approved by the European Medicines Agency (EMA) for treating children with ASD [10]. The regulation of immediate-release melatonin products <2 mg varies across countries in the European Union [11], while these products are considered dietary supplements in other countries, where they lack regulation, quality assurance, and safety parameters.

Prolonged-release melatonin has been evaluated in different patient populations, proving its efficacy and safety [12-14]. Children with ASD or Smith-Magenis syndrome were treated with the EMA-approved PedPRM (colorless and odorless mini tablets) in a double-blind, placebo-controlled, international phase 3 trial, where it demonstrated long-term safety and efficacy by significantly increasing total sleep time, improving the time slept continuously, and decreasing sleep latency [15-17]. Besides improving sleep, treatment with PedPRM resulted in a significant improvement in the child’s externalizing behavior, which was also associated with improved parent well-being [18].

Other than the phase 3 clinical trial results for PedPRM, real-world data on patients with ASD and sleep disturbances are scarce. Here, we describe the effectiveness of PedPRM in real clinical practice in a heterogeneous, highly complex, and refractory patient population in Spain.

In this study, we report the effectiveness of PedPRM, which led to significant improvements in sleep despite the small sample size of a highly comorbid and heavily medicated patient population that was refractory to various treatments for sleep disturbances. The patients’ refractoriness highlights the need for effective therapy to treat insomnia in children with ASD and underscores the remarkable improvement in sleep achieved with PedPRM in real clinical practice. There was a significant improvement in daytime sleepiness, as determined by the PDSS, with starting scores close to the pathological cut-off. We did not observe an improvement in total sleep time, which could be attributed to missing values or to most patients already sleeping at the acceptable/recommended level for their age when the study started. However, the number of nocturnal awakenings and the sleep latency both declined significantly when using PedPRM. Although patients did not sleep for a longer time overall, a reduction in sleep latency and sleep fragmentation is important for improving quality of sleep and sleep continuity, which have been associated with fewer behavioral difficulties during the day [23]. Notably, most patients continued receiving PedPRM after the 6-month evaluation, given the good results obtained.

The etiology of sleep disorders in children with ASD is multifactorial and includes anxiety, depression, abnormal melatonin synthesis, epilepsy, changes in routine, and hypersensitivity to stimuli [6,24]. Studies have reported lower levels of endogenous melatonin in children with ASD [25,26] along with mutations in melatonin pathways [24], supporting the use of PedPRM in these patients. In a phase 3 trial of children with ASD, meaningful results were obtained using PedPRM [16]; however, the patient population was less comorbid than the one included in our study. In the quoted study, 12.8% of patients had epilepsy [16], compared with 73.9% in ours. Some of the comorbidities presented by the patients reported here are particularly serious, such as epileptic encephalopathy in 30.4% of cases [15]. Furthermore, some patients were taking concomitant medication, including some that were prohibited in the above-mentioned phase 3 trial (alimemazine, clonazepam). Patients in our study were refractory to many medications. All patients had used immediate-release melatonin (compared with 65.6% in the phase 3 trial), and 87.0% of them had also used immediate-release melatonin plus tryptophan before enrolment. Moreover, while all patients in the phase 3 trial began treatment at a dose of 2 mg, some of the patients in our study began with higher doses. Overall, achieving a meaningful improvement in sleep in the population included in our study is very encouraging for the use of PedPRM in clinical practice, given its limited interaction with other drugs and its good tolerability [9].

Expanded access programs (also known as early access, preapproval access, or compassionate use) enable patients to receive therapy that would otherwise not be available to them [27]. These programs are also valuable sources of real-world evidence for the efficacy and safety of treatments [28] and can sometimes support the clinical efficacy profile of a drug [27]. In our study, this approach enabled us to improve sleep disturbances in a challenging patient population. PedPRM is currently not reimbursed by the Spanish National Health System. We hope that our findings will help change clinical practice and the perception of PedPRM by showing that meaningful results can be achieved, even in complicated cases. Some patients in our study started PedPRM treatment at a dose higher than that recommended; nevertheless, we suggest following the treatment algorithm proposed by Schroder et al. [8], where the starting dose of PedPRM is increased from 2 to 10 mg if feasible and necessary, regardless of the dose of immediate-release melatonin used previously and re-evaluating the dose every 6 months.

There is scarce data on the real-world use of PedPRM in patients with ASD. The main strength of this study lies in the observed effectiveness of PedPRM in improving sleep in a heavily medicated, heterogeneous population with serious comorbidities such as epilepsy. In addition, we used several tools to evaluate sleep, including the SDSC, one of the most widely used and methodologically valid tools [29]. This study had several limitations. First, the study sample size was small, partially related to the unapproved status of PedPRM in Spain at the time of this study, as this treatment needed to be approved by the hospital pharmacy. Second, the patient population was highly heterogeneous and heavily medicated with antiepileptic drugs, which might have affected sleep. Third, missing values due to loss to follow-up may explain the lack of significant differences in some variables, such as total sleep time (Table 3). Missing values also resulted in some parameters not being evaluated in all patients. Missing information and the small sample size precluded comparing efficacy of PedPRM in patients with controlled versus uncontrolled seizures. Fourth, there was no control group, and sleep was not evaluated with an objective instrument such as an actigraphy monitor [30]; however, this device was refused by most patients with ASD and insomnia in a previous study and, thus, may not be a viable option [16]. Finally, the starting dose of PedPRM given to patients did not follow current recommendations, given they had all been previously treated with high-dose immediate-release melatonin and other treatments for insomnia. Of note, this study was carried out before the recommendations by Schroder et al. [8] were published.

In conclusion, the current study demonstrated in real clinical practice that PedPRM resulted in improved sleep in a complex pediatric patient group with ASD and severe comorbidities, such as epileptic encephalopathy, for whom previous pharmacological hypnotics had been ineffective, even at very high doses. We hope these findings encourage the use of PedPRM in complicated cases that were not represented in the phase 3 clinical trial where PedPRM was evaluated.