Objective: To explore the methods for generating evidence on health outcomes in children with rare diseases. Methods: The data from 30 clinical trials on rare neurological diseases in children from January 2008 to December 2018 were collected and analyzed.Statistical analysis was conducted on the relationship between the study sponsor and the study center, the number of participants and the prevalence rate. Results: Thirty studies involved 6 types of diseases, including 14 kinds of diseases.(1) All global multicenter studies (14 items) were initiated by pharmaceutical companies, whereas most of single-center studies (6/7 kinds, 86%) and multiple centers within one country(7/9 kinds, 78%) were initiated by investigators.There was a significant correlation between the research center and the research sponsor(P<0.001). (2) Most of the drugs studied were selected based on previous clinical trials (9/30 items) and animal experiments (9/30 items). (3) The median number of participants included 39 cases (10-215 cases), and 60%(18/30 items) of the studies was fewer than 50 cases.(4) Study design: 53%(16/30 items) of studies were randomized controlled studies, 33%(10/30 items) studies were open-label single-arm studies, and 14% (4/30 items) were randomized cross-over trials.Seventy-five percent(12/16 items)of randomized controlled studies were initiated by pharmaceutical companies, 50%(5/10 items) open-label single-arm studies and all randomized cross-over trials were initiated by investigators.There was a statistical correlation between the study sponsors and the study design method (χ²=7.602, P=0.022). (5) Outcome index: Scale score was used as the primary outcome in half of studies.Other studies used symptom improvement or pathological changes. Conclusions Clinical trials in rare diseases enrolled fewer participants than that in non-rare diseases, and the study design method was relatively simple.Therefore, it is necessary to further improve the level of evidence of clinical research of rare diseases through global and multi-center recruitment, initiation of pharmaceutical companies and improving the study design method.

Objective:To analyze the clinical features and genotypes of mitochondrial complex Ⅰ deficiency due to NDUFAF5 gene mutations.Methods:Clinical data of 2 cases with mitochondrial complex Ⅰ deficiency due to NDUFAF5 gene mutations admitted in the Department of Pediatrics, Peking University First Hospital from February 2015 to July 2018 were retrospectively reviewed and followed up.Reported cases of mitochondrial complex Ⅰ deficiency due to NDUFAF5 gene mutations were searched in online databases, including the PubMed, Wanfang, Chinese Journal Full-Text Database and VIP database from January 1975 to February 2020 with " NDUFAF5" as the key word.Through literature review, clinical features and genotypes of mitochondrial complex Ⅰ deficiency due to NDUFAF5 gene mutations were summarized.Results:Case 1 showed mentor and mental regression after infection at the age of 1 year and 4 months.The condition of case 1 remained stable at the age of 5 year and 6 months at the last follow-up.Brain magnetic resonance imaging (MRI) showed multiple lesions in the white matter of the frontal and parieto-occipital lobes, basal ganglia, thalamus, cerebellum, brain stem and corpus callosum.Case 2 showed rapidly bilateral visual impairment at the age of 7 years and 4 months.The patient′s vision moderately recovered at the age of 8 years and 8 months.Brain MRI showed midbrain, periaqueductal gray, medulla oblongata and putamen lesions.Spinal MRI showed continuous lesions in the cervical cord 1-4.Genetic test showed NDUFAF5 gene c. 764C>T (p.Ala255Val) and c. 508C>T (p.Arg170Trp), homozygous c. 836T>G (p.Met279Arg) mutations in case 1 and case 2 respectively.Through online searching, 6 reports involving 14 cases were retrieved.The most common clinical phenotype was Leigh syndrome.Two cases had disease onset during the neonatal period, and their disease progressed rapidly and died within 1 year old.Eleven cases had onset during the infantile period, and 72.7% (8/11 cases) of them had a normal development.The common initial symptoms were mental or motor regression, feeding difficulty and dystonia.Seventy-two point seven percent (8/11 cases) had acute/subacute onset after infection, showing paroxysmal deterioration, and died in infancy or childhood.One patient developed dystonia in childhood and visual impairment in adulthood.Conclusions:The onset age ranged from neonatal period to childhood in patients with NDUFAF5 gene mutations, and their clinical phenotypes vary a lot.The main clinical phenotype is Leigh syndrome.Disease onset during the infantile period is frequent, and mostly presents paroxysmal deterioration after infection, while disease onset in childhood is rare.