Prevalence of Pediatric Neuromuscular Disorders in pediatric patients: a consensus at last?
Our overall prevalence is based on a large cohort of patients under 18 years (838 in all and 775 still alive) diagnosed with NMD and followed-up in three experts centers between 2001 and 2022 in Southwest region. Even if this study only includes regional data, our data are current and of good quality and our prevalence could not greatly differ from French national data. We found a regional prevalence of overall NMD patients under 18 years old of 37.9 for 100,000 inhabitants. This prevalence is comparable to those found by Woodcock et al. (2016), which is one of the few pediatric prevalence studies. In their study done in Yorkshire, authors found an overall prevalence of NMD conditions of 36.9 per 100,000 in a population under 16 years old. NMD included in their studied population is very close to ours, so the similarity in the results adds to their validity.
Few other researches addressed the special issue of prevalence in pediatric population of NMD or as part of a larger study on general population (Theadom et al., 2019; Pagola-Lorz et al., 2019; Rose et al., 2019). Our results differ, but the reliabilities of the pediatric prevalence announced in these studies is questionable. Rose et al. (2019) conducted a very large cohort study based on the population of Ontario (Canada) where adults and children with NMD were rigorously identified using health administrative databases. 27,823 children were eligible for 2014' cohort creation establishing an annual prevalence of 75.9. However, authors accounted ‘cerebral palsy’ (24.1) and ‘spina bifida’ (13.6) in NMD diagnoses, which is quite uncommon in such NMD prevalence studies. Excluding these two categories, annual prevalence in NMD children falls 38.2 for children, very close to ours.
Pagola-Lorz et al. (2019) and Theadom et al. (2019) report a pediatric prevalence of NMD of 21.87 and 28.11 respectively, but both excluding SMA and regarding data under 14 years, while prevalence greatly increases with age in both.
Our pediatric prevalence is also quite different of those identified in Muller et al. study routinely cited in support of NMD prevalence. The total NMD pediatric prevalence in their study (57.8 per 100.000) was higher than the prevalence of NMD related by our study and studies previously cited. But authors themselves added that the prevalence found in their study greatly higher than other, including higher than proposed in epidemiological study that combined populations from different parts of the world (Emery, 1991 for example). If their finding could be due to a small number of children under 18 years old (55), it may also reflect ethnic difference according to the authors. This explanation seems confirmed by the results of several studies (Theadom et al., 2019; Woodcook et al., 2016) that highlighted some considerable variation in prevalence by ethnic groups.
To date, an NMD’ pediatric prevalence about 38/100000 may reflect a consensus among different countries. However, our work highlights an NMD prevalence fluctuation between studies realized over the last decades. This suggest that significant variation between methodologies or datasets used greatly impact prevalence estimation. The NMD included, the age groups studied, the ethnicities concerned, varied to one study from another. Studies comparisons are therefore problematic. Another point that hindered the possibility of comparison between studies, is the period studied. Innovative therapies are in constant evolution, and compared prevalence before or at the beginning of their availabilities is not rigorous since they progressively increased the useful life of NMD patients, especially for young children. Concerning pediatric, very few researches actually addressed the special issue of prevalence in pediatric population of NMD, intrinsically or as part of a larger study on general population. But the arrival of new very expensive therapies forces us to rigorously estimates the number of patients involved at a territorial, national or world level, to better inform healthcare policy on the burden of such care on the healthcare system. We therefore urgently need rigorous prevalence studies, with a consistent and standardized methodology between countries, applicable and adopted by all. The same observation was previously made on numerous occasions in studies or review for many years, which express the need to greater consistency in the conduct of NMD epidemiological studies to allow and ensure comparisons to be made between studies. However, nothing happens despite concrete and practical proposals (Theadom et al., 2014). Accordingly, prevalence studies guidelines should be discussed and adopted by consensus in world conference on NMD. It is essential to homogenize inclusion or exclusion criteria, selection of conditions that have been included or excluded from the definition of NMD, the use of more inclusive or rather the opposite more limited definition, to clearly defined some range-ages of interest used from study to study, to sought and noted the ethnicity of patients, to specified the datasets used (research or clinical or multiple data sources), etc. Only in this way will prevalence rates should be compared across studies.
Prevalence temporal trends
Two studies in the recent past have underlined the number of people with MNM has been steadily rising (Carey et al., 2021; Rose et al., 2019). Rose et al. (2019) used health administrative databases to describe trends in incidence, prevalence, and mortality of adults and children with NMD on a population-based (Ontario, Canada) cohort study (2003 to 2014). Authors observed a rising prevalence of NMD over time among both adults and children (2003 to 2014). Carey et al (2021) used the Clinical Practice Research Datalink, a primary care database in the UK, to estimate trends in the recording of neuromuscular disease in UK primary care between 2000–2019 based on incidence and prevalence rates in each year. Authors observed overall prevalence grew by 63% since 2000 with temporal trends showing the number of NMD patients is steadily increasing year by year. For Carey et al. (2021) as for Rose et al. (2019), new cases cannot solely explain increasing prevalence year by year, as incidence remained constant, but that it may be due to better recording.
Our data, which run from 2001 to 2021, support this assumption. We can see a continuous, steady growth in rate of prevalence from 2001 to 2017 (Fig. 6). In France, this period corresponds to the historical implanting of centers of rare diseases and to the creation and progressive implementation of national research database, CEMARA (Messiaen et al., 2008) at first and then BAMARA (Jannot et al., 2021). 2017 is a crucial point in France with the introduction of BAMARA, its deployment in all rare diseases’ centers, and its used making systematic and compulsory as part as third rare diseases plan, leading de facto to better and reliable healthcare data recording. Since 2017, our data are therefore more reliable. And watching exclusively on research' data specifically covering these last six years, we don’t find an increasing of prevalence of NMD per years but rather a clear stabilization in the number of NMM children followed expert care centers has been recognized.
Increasing pointed by Rose et al. (2019) and Carey et al. (2021), as by our own data from 2017, might therefore be an artefact being but the reflection of a data filling effect. Our finding concurs with the Carey et al. (2021) assumption, in believing that this increasing may partially be due to better recording.
In addition, most studies that contain substantive pediatric data, found that prevalence increases with age across all diagnoses except SMA. Rose et al. (2019) found that childhood disease prevalence increased by 10% per year. For authors, the largest increase was in children 0 to 5 years. Theadom et al. (2019) that found that prevalence remained relatively stable across the lifespan following 5 years of age have enhanced this finding. Our prevalence increases with age, to 12.7/100000 in the age group 0–5 years, from 53 in the 15-18-years age group (Table 2). As was already demonstrated, we show a very significant gap forms at the age of 5, with more than a two-fold increase prevalence after 5. Then, the prevalence increases progressively but we note a stabilized trend after 10 years. Incidence and prevalence data years by years in pediatric population will be required, probably by diagnosis, in order to better understand the course of NMD throughout childhood.
Focus on DMD, BMD, CMT1 and SMA
The most common NMD in our study was dystrophinopathies (17.6%: 13.2% Duchenne and 4.4% Becker), followed by CMT1 (14.5%), SMA (11%) and DM1 (8.5%). These four diagnoses made more than half of the overall prevalence of our NMD population.
Few NMD prevalence study exists on pediatric population and distribution of NMD diseases differs between children and adults. However, our results are quite congruent with previous studies on pediatric prevalence of NMD. Woodcock et al. (2016) found Dystrophin-related NMD as the most prevalent condition of their population of 261 NMD children, 16.9 per 100,000 (61; 23%), followed by CMT1 (31; 12%), congenital myopathies (29; 11%) and SMA (27; 10%). In Thongsing et al. (2020), in a quite as large pediatric NMD population (217), the most common inherited NMD were the Dystrophinopathies, including Duchenne / Becker muscular dystrophy (58; 27%), followed by SMA (25; 11.5%) and Hereditary Motor Sensory Neuropathy (CMT1) (16; 7%). Our study therefore strengthens its results on a three-time larger sample.
Focus DMD/BMD
DMD/BMD occurs in males, but in order to calculate the scope of the burden for society, most studies report prevalence estimates in relation to the general population. We did the same and our estimate is rather comparable to most previous studies and to systematic review and meta-analysis of Crisafulli et al. (2020), Theadom et al. (2014) and Mah et al. (2014). Crisafulli et al. (2020) found a DMD prevalence of 2.8 cases (95% CI: 1.6–4.6) per 100,000 in the general population on about 40 studies reporting the global epidemiology of DMD. Theadom et al. (2014) found a prevalence for DMD of 1.7–4.2 per 100,000 and for BMD of 0.4–3.6 per 100,000 based on studies classified as having a low risk of bias (15/38). The systematic review and meta-analysis of Mah et al. (2014) analyzed the prevalence for DMD and BMD but in relation to male population only. Authors found a prevalence of DMD at 4.78 and BMD of 1.53 (95% CI 1.94–11.81) per 100.000 males based on 31 studies. Comparing findings with two alternative methods is a challenge, but Crisafulli et al. (2020), Theadom et al. (2014) and Mah et al. (2014) estimates seems relatively similar, and our prevalence appears little higher -at least for DMD. However, these reviews considered prevalence in all age groups and not in pediatric groups. Difference thus could be probably due to the early mortality of DMD patients. Despite improvement of survival, few affected individuals survive beyond the third decade (Passamano et al., 2012) with a median survival of 24 years (Rall & Grimm, 2012). Because death occurs in early adulthood, a higher pediatric prevalence seems therefore logical.
Parents can identify the first symptoms of DMD early, when physical ability in their children diverges markedly from that of their peers around 2–3 years (Mercuri et al., 2019). Significant and visible general motor delays (gait problems, delay in walking, a waddling gate, difficulties with climbing stairs, and frequent falls…), then a little later learning difficulty, and speech problems, are indeed perceptible from early development. In our DMD cohort, ours results are congruent with the recent study of D'Amico et al. (2017) that have reported mean age at diagnosis was 41 months (range 0.3–135 months), 10 months (range 10 days to 80 months) before the first suspicions (mean age 31 months; range 0–95 months). Authors note that it’s about one year less than the age reported in Bushby et al, (2010, probably related according to them to the fact that in Italy blood tests including transaminases and CK are routinely requested by pediatricians when motor delays is present, as in France (Verloes et al., 2012). We agree therefore the authors to say that a CK test screening should be performed in early infancy in order to reduce the delay of diagnosis.
Regarding BMD, we observed a median age of 6.0 years at first signs (IQR = 3.6–9.5) whereas the median age at diagnosis was 7.5 years (IQR = 5.0–10.2). Few studies focused on natural history of BMD when the onset is in childhood, which does not allow comparison, but our data seems congruent with clinical practice. BMD is indeed less severe and has a milder clinical course than DMD. The onset of symptoms is usually later than in DMD. The clinical heterogeneity of BMD is extreme, and the age of onset varies widely (Angelini et al. 2019). The spectrum of clinical presentations ranges from asymptomatic with screening via a liver test to a loss of ambulation occurring in the teenage years.
Focus CMT1
We found a pediatric prevalence of CMT1 at 6.2 (5.1–7.3) per 100.000, without analyzable data between 0 to 5 years and with first reliable epidemiological data after 5 years. Epidemiological studies of CMT disease are scarce, and in the absence of data collection programs, longitudinal or natural history studies, knowledge of CMT pediatric epidemiology is even more limited. Patients have a long history of symptoms before the diagnosis, that come in late in life, with a mean age estimated of 31.8 years (Gudmundsson et al., 2010), often caused by discreet symptomatology, slow progression rate, and insidious onset in the first decades of life. The average delay of clinical diagnosis, even in families known to have CMT, is more than 10 years (Jani-Acsadi et al., 2015). In a cohort of 39 patients with infantile presentation, the mean age at diagnosis was 8.5 years (Ounpuu et al., 2013). Mean age at diagnosis and mean age for first symptom in the CMT1 subtype for pediatric population is not available to our knowledge, but for our part, we were able to estimate a median age of first symptoms of 4.3 years (IQR = 2.0-8.1) and a median age at diagnosis at 7 years (IQR = 4.0-10.3).
With regard to prevalence, Ma et al. (2023) describe the distribution of CMT disease among the worldwide population in a relevant meta-analysis. Authors examine the prevalence of CMT for the general population, as well as the subgroups (age, gender, region, and disease subtypes). 31 included studies are detailed in their meta-analysis but only 2 old studies (1983 and 2000) focused on non-adult populations, and some few conducted on the all-age populations (comprising children). CMT prevalence is logically higher in older than in younger age groups. Extracted from their meta-analysis, only Carey et al. (2021), Theadom et al. (2019) and Mladenovic et al. (2011) provided recent pediatric epidemiological data (post 2010). However, although both Carey et al. (2021) and Theadom et al. (2019) indicated the estimates prevalence for different subtypes, they did not discriminate CMT in subtypes for pediatric population. The pediatric prevalence for CMT1 only is not available. Theadom et al. (2019) found a pediatric prevalence for all CMT at 9.1 (6.1–13.5) quite similar to Carey et al. (2021) that found a pediatric prevalence for all CMT at 8.9. In literature, CMT1 has been reported to be the most common CMT type, accounting for between 37.6% and 84.0% of cases (Barreto et al. 2016). This estimation brings up the CMT1 pediatric population of both studies to 3.4 to 7.6, in line with our finding and those of Mladenovic et al. (2011). Indeed, in Mladenovic et al. (2011) study, as in ours, CMT1 prevalence under 5 years old is unavailable and authors found a pediatric prevalence for CMT1 at 5/100000 (1.6–11.6) under 14 years old. Overall, based on previous studies and our data, we can conclude that pediatric prevalence of CMT1 could be endorsed about 5 per 100.000.
Focus SMA
The regional prevalence of patients affected by SMA followed-up in expert centers was 3.2 [2.5–4.1]. Only a few estimations studies have been performed to assess the prevalence of SMA and most of these have been conducted before 2000. When examining all types of SMA together in these studies, a prevalence of around 1–2 per 100,000 persons is observed (Verhaart et al., 2017) for most country. A more recent study was available, conducted by direct contact with two genetic laboratories across Europe (Verhaart et al., 2017b). In this study, even if there was considerable inter-country variability, SMA prevalence ranged respectively from 0.01 to 2.43 per 100,000 (TREATNMD Global SMA Patient Registry) and 0.00 to 4.11 per 100,000 (Care and Trial Sites Registry CTSR). However, both registries only included patients before 2014 and the launch of innovative therapies that increases significantly the survival of patients with SMA. With a growing number of therapies being developed since 2017, the natural history of SMA has indeed changed. Moreover, pediatric prevalence and overall prevalence (lifespan, children and adults) cannot be easily compared due to high early death for a substantial number of patients. There is therefore an increasing need for others new pediatric SMA incidence and prevalence estimations.
The majority of SMA children in our cohort were classified as SMA type I (79/179 or 44%) followed by 59 with SMA2 (33%), 29 with SMA3 (18%), 10 with other SMA. This is in good agreement with the percentages of patient classified into different types of SMA in the Cure SMA database (one of the largest patient-reported data repositories on SMA patients worldwide (Belter et al., 2018)) that found: SMA1 39%; SMA2 31%; SMA3 19% (Sun et al., 2023).
The percentages of females of SMA (56.4%) were higher than males for overall SMA cohort (male-to-female sex ratio 0.77) and for all SMA types (male-to-female sex ratio 0.76 for SMA1 and 0.45 for SMA3 and 0.9 for SMA2). This finding is interesting since no consensus has been reached for this question. An old study from Pearn (1978) reported a male to female ratio of 2.0 in SMA1, but a sex ratio in male disadvantage is commonly admitted. However, these have not been consistent, and several studies have reported no sex differences in sex ratio. This question could be addressed, especially because sex vulnerability in SMA is identified with some studies that have indicated the infantile form of SMA is more severe in males (Sun et al. 2023).
The first medical visit for SMA patients in the reference center came early with a median age of 1.4 (IQR = 0.4–8.2), which is consistent with the disease evolves and its gravity. In the first year of life, 72 patients (53%) were diagnosed.
We observed for overall SMA first signs and age at diagnosis at a median age of 0.8 year, corresponding to a rapid diagnostic. By SMA subtypes, the gap between first signs and diagnosis is quite similar for SMA1 and SMA2 (few weeks) and increased for SMA3 (1.5 years between emerging of first symptoms and diagnosis).
In our study, survival for SMA1 children at 1, 2, 4 and 8 years was 36%, 29%, 27% and 27%, respectively. The survival probabilities of SMA2 patients at 1, 2, 4 and 8 years was 100%, 100%, 97% and 93%, while SMA3 patients generally have normal life expectancy. These findings are similar with those from Farrar et al. (2013) or Chung et al. (2004) and concordant with clinical experience. We just observed a slight difference for the clinical course and death for SMA1, which probably reflect advances in medical care, and improvement in life expectancy.