In this retrospective study, the prevalence of feeding problems in children with genetically confirmed NS was assessed. Molecular testing identifies mutations in PTPN11 in about 50%–60%, SOS1 in about 15%, RAF1 in 3–17% and the other genes in 5% or less of genetically proven patients with NS and NL syndrome [14]. In our population, we found the same distribution of type of mutation. The prevalence of feeding problems in the first year was 52% and more than half of our patients were tube-dependent. This is consistent with previous studies on this subject. Feeding problems in association with Noonan syndrome were already reported in 1992. In this study, more than 76% of patients with clinically diagnosed NS experienced feeding problems in infancy, of which almost one-third was tube-dependent [10]. In 1999, feeding problems were described in 64% of young children with clinically diagnosed NS [15]. All these children required tube feeding. Shaw et al. published in 2007 that 65% of children with clinically diagnosed NS experienced feeding problems in the first months of life, of which 36% were tube-dependent [11]. The most recent study on this subject, from Croonen et al., described feeding problems in 56% of children younger than 1 year old with clinically and genetically proven Noonan syndrome [16]. More than half of these children (59%) were tube-dependent. Our results are consistent with this recent paper. There was no difference in the prevalence of feeding problems in infancy in patients with a PTPN11 or SOS1 mutation. However, the prevalence of early onset feeding problems in patients with a gene mutation with low prevalence was much higher with the same documented presumed or proven cause.
The possible course of feeding problems was only mentioned in the study of Shah et al. [15]. They mentioned that the feeding problems remarkable improved after the age of 3 to 4. In our study, we found a major improvement of the feeding problems between the age of 1 to 2, with only a minority having still feeding problems after the age of 2. In patients with a SOS1 mutation the feeding problems improved even significantly faster. When feeding problems persisted, children most oftenly stayed tube-dependent. The documented presumed or proven cause of the feeding problems was most often gastro-oesophageal reflux disease (68%). Gastro-oesophageal reflux disease was documented in 44% of children with feeding problems by Shah et al., as well as delayed gastric emptying [15]. At the present time, the diagnosis of gastroesophageal reflux disease can be made based on clinical symptoms. An upper gastro-intestinal study can be used for evaluating other anatomic causes of vomiting. Upper endoscopy is only considered for children with concerning symptoms, persistent symptoms despite treatment, and relapse of symptoms after treatment [17]. Delayed gastric emptying may exacerbate any tendency to gastroesophageal reflux disease [15]. Gastric emptying studies may play a role in the diagnosis of delayed gastric emptying [18]. We think, and this was also postulated by Shah et al. [15], that the cause of the feeding problems is a delayed gastrointestinal motor development and that this improves with age.
We could not study growth as a consequence of the feeding problems, due to inappropriate data. However, in an earlier study with another design, we showed that children with Noonan syndrome and feeding problems were clearly lighter and shorter and had a clinically relevant lower weight and length SDS at the age of 1 than children without feeding problems [16].
Remarkable was the fact that almost 1 out of 7 children developed feeding problems at an older age. This was frequently due to postoperative circumstances, infections or co-morbidities, and not directly due to the causes of early-onset feeding problems. However, it is known that infections, or (operations for) heart diseases and cancer are more common in children with NS [19,20,21]. So having Noonan syndrome could be an indirect factor of late onset feeding problems and children with NS might be prone to feeding problems that can be triggered by, for example, an infection. This is illustrated by the case report that delayed gastric emptying may even present for the first time during adulthood [22]. No children were diagnosed with celiac disease, although Quaio et al. notify that physicians should be alerted to the possibility of autoimmune diseases, including celiac disease, in patients with NS [23].
This study has some limitations. Due to the retrospective design, many patients had to be excluded. However, the included patients had the same distribution of type of mutation as in the literature, and the recorded prevalence was in the same range as in the literature. So, we think, the recorded prevalence data may be reliable. Another problem due to the retrospective design, is that follow-up data were limited and not always straightforward. Moreover, symptoms of the existing feeding problems were not always clearly documented, and the cause of the feeding problems was often not studied extensively. The scoring system used in the literature could not be used due to the retrospective design [10, 16]. For this reason, we used the validated Functional Oral Intake Scale [12, 13]. This scale may underestimate the prevalence of (behavioural) feeding problems, thus underestimating the prevalence. Moreover, there was no information on psychiatric comorbidity and/or behavioural problems in our retrospective data in these young children.
There is no literature on psychiatric/behavioural problems in young children (< 4–6 years) with Noonan syndrome. However, it is known that feeding problems in infants and young children are caused by different issues like medical and/or developmental history, problems with motor control and function, problems with swallowing and behaviour, that all may play a role and may interact with each other [24]. Nevertheless, this study is the first to combine prevalence of feeding problems, the course and presumed or proven causes of feeding problems and the association with the genetic mutation. Another limitation is the fact that, using the ClinGen gene curation framework, it can be questioned if the used diagnosis is correct [25]. So is an HRAS-mutation definitively associated with Costello syndrome, and is a BRAF-mutation moderately associated with NS and definitively associated with cardiofaciocutaneous syndrome. However, as these children are collected in the group other gene mutations, the results still may reflect the prevalence of feeding problems in patients with NS.