Gastrointestinal symptoms and channelopathy-associated epilepsy

Abstract

Objective:

To determine the prevalence of and identify factors associated with gastrointestinal (GI) symptoms among patients with channelopathy-associated developmental and epileptic encephalopathy (DEE).

Study design:

Parents of 168 patients with DEEs linked to SCN1A (N=59), KCNB1 (N=31), or KCNQ2 (N=78) completed online CLIRINX© surveys about their children’s GI symptoms. Analysis examined prevalence, frequency, and severity of GI symptoms, as well as DEE type, functional mobility, feeding difficulties, ketogenic diet, anti-seizure medication, autism spectrum disorder (ASD), and seizures. Statistical analyses included chi-square tests, Wilcoxon rank-sum analyses, and multiple logistic regression.

Results:

GI symptoms were reported in 92/168 (55%) patients among whom 63/86 (73%) reported daily or weekly symptoms, 29/92 (32%) had frequent or serious discomfort, and 13/91 (14%) had frequent or serious appetite disturbances as a result. The prevalence of GI symptoms varied across DEE cohorts with 44% of SCN1A-DEE patients, 35% of KCNB1-DEE patients, and 71% of KCNQ2-DEE patients reporting GI symptoms in the previous month. After adjustment for DEE type, current use of ketogenic diet (6% reported), and gastrostomy tube (13% reported) were both associated with GI symptoms in a statistically, but not clinically significant manner (P < 0.05). Patient age, functional mobility, feeding difficulties, ASD, and seizures were not clearly associated with GI symptoms. Overall, no individual anti-seizure medication was significantly associated with GI symptoms across all DEE cohorts.

Conclusions:

GI symptoms are common and frequently severe in DEE patients.

Introduction

Patients with developmental and epileptic encephalopathies (DEEs) often exhibit a wide range of impairments, which can complicate treatment and reduce quality of life (1). One often distressing concern that is repeatedly found in clinical reports of patients with DEE, including those with variants in genes encoding voltage-gated ion channels, is gastrointestinal (GI) dysfunction, particularly constipation and dysmotility (1–13). These anecdotal reports, however, have not been systematically explored or characterized.

Multiple factors may be linked to the development of GI symptoms in the DEE patient population. Physical inactivity contributes to constipation in the general population (14, 15), and many patients with DEE often have limited mobility (16). The ketogenic diet, used as a treatment for intractable seizures in many DEE patients, as well as the use of enteral nutrition in DEE patients with feeding difficulties, are potential causes of GI symptoms (17–19). Many anti-seizure medications are associated with adverse GI side-effects (20). Finally, many DEE patients have also been diagnosed with autism spectrum disorder (ASD) (21), which is frequently accompanied with increased reports of GI symptoms, with constipation being most common (22, 23). The etiology of the GI disturbances observed in ASD patients is not known; decreased physical mobility, diet, and side-effects of medications have been postulated as potential causes (22, 23). Conversely, there is anecdotal clinical evidence as well as animal studies suggesting that constipation can induce seizure activity (24).

Ion channel gene-linked DEE variants may play direct roles in GI dysfunction. Voltage-gated sodium (Na_v) and potassium (K_v) channels are expressed in the mammalian GI tract, where they regulate the activity of enteric neurons and other excitable cells, thereby contributing to proper GI motility and fecal expulsion (25–38). Thus, DEE-linked genetic variants may potentially directly affect GI function by causing aberrant GI cell excitability.

There is an unmet clinical need to understand the mechanism(s) of GI symptoms in DEE. As a first step toward this goal, we used responses from a parent survey that included a large number of children with DEEs linked to variants in SCN1A, KCNB1, or KCNQ2 (which encode for the voltage-gated ion channels, Na_v1.1, K_v2.1, and K_v7.2, respectively) to uncover potential associations between the presentation of GI symptoms and known risk factors for GI dysfunction: mobility, diet, medication, and ASD. An improved understanding of the factors and biological mechanism(s) underlying the recurring GI symptoms observed in DEE patients may lead to the development of novel treatments that address this disruptive comorbidity and improve quality of life for patients and caregivers.

Methods

Survey design and participation:

An online parent-reported survey was performed using the CLIRINX® platform (Dublin, Ireland) and disseminated to families through several organizations, including the Dravet Syndrome Foundation, the KCNB1 family group, and the KCNQ2 Cure Alliance. Caregivers of DEE patients of all ages were invited to participate. All survey data were collected through parent reports from June 2018 through February 2020. Parents were invited to upload their children’s genetic testing reports; however, we based inclusion of a child in this study on the parent’s affirmation that the child had a causative variant in one of the three ion channel genes of interest: SCN1A, KCNB1, or KCNQ2. Full methods for the survey have been previously described (16). The following provides a synopsis.

Gastrointestinal (GI) symptoms:

Parents were asked to indicate whether their child had experienced specific GI symptoms—constipation, gut dysmotility or diarrhea—over the past month. They could also write in additional symptoms. As virtually all the symptoms endorsed or reported were either constipation or dysmotility, and these two symptoms overlapped in all but a few cases, we simplified our primary outcome to the reporting of “constipation, dysmotility, or both” versus “neither”. For children with GI symptoms in the previous month, parents reported typical frequency (“daily,” “weekly,” “monthly,” or “less than once per month”) as well as whether the symptoms caused either discomfort or affected the child’s appetite (“no,” “yes – occasionally but not a serious concern,” or “yes – common occurrence or serious concern”). Parents also reported if they had used any specific methods to alleviate the GI symptoms (supplements, enemas, medications).

Potential risk factors associated with GI symptoms:

Parents provided information concerning functional mobility (for children ≥ 2 years), dietary therapies, feeding tube usage, current anti-seizure medications, and diagnosis or features of autism.

For mobility, we used the Gross Motor Functional Classification System (GMFCS) (39). For analytic purposes, the GMFCS scale was condensed into three levels: “dependent” (scores of 4 or 5), walks but “requires aids” (score of 3), and “independent” (scores of 1 or 2).

Parents were asked if their child had a feeding tube and, if so, whether their child was exclusively tube-fed or partially tube-fed. For children who ate by mouth (including those partially tube fed), parents indicated how often a child had difficulty chewing or swallowing. Choices were: “multiple times per meal,” “usually once per meal,” “not every meal but at least once on most days,” “not daily but usually a few times per week,” “less than once per week but every now and then,” or “almost never.” These responses were condensed into the categories: “typically at least once per meal,” “typically more than once per day,” “typically more than once per week,” or “rarely or never”. Parents were also asked whether their child was currently being treated with a classic ketogenic, modified Atkins, or other special diet.

Parents reported if their child had an ASD diagnosis or presented with features of ASD. Parents reported on the occurrence of seizures in the previous six months and whether bowel movements were considered by the parents to be a trigger for seizures.

Finally, parents reported the child’s current use of 35 specific anti-seizure medications. For overall analysis of all three DEE cohorts, only medications being taken by 10 or more patients were considered. For individual analysis of each DEE cohort, only medications that were being taken by five or more patients were considered.

Data analysis:

Statistical analyses were performed in SAS© Methods included chi-square tests, Wilcoxon rank-sum analyses, and multiple logistic regression. Relative risks were used to quantify the strength of associations when appropriate. Results were considered significant at P < 0.05.

IRB approval and informed consent:

All procedures for this study were approved by the Lurie Children’s Hospital IRB. Parents provided informed consent electronically through the CLIRINX® platform.

Results

Participants:

Of parents who enrolled in the study, 168 affirmed their child’s pathogenic variant in one of the three ion channel genes of interest, SCN1A (N=59), KCNB1 (N=31), or KCNQ2 (N=78), and completed the survey portions about GI symptoms. Median patient age at the time of survey participation was 6.3 years (interquartile range: 3.3 to 10.3 years), and 55% of affected children were female (Table 1). The overall age distribution of DEE patients was: <5 (43%), 5–9 (30%), 10–15 (15%), and ≥ 16 years (12%). The children in KCNQ2-DEE group had a substantially younger age than those in the other two groups (P < 0.001). The sex distribution was similar across groups (P = 0.97; Table 1).

Table I.

Demographic characteristics of DEE cohorts

	N	Sex N (% female)	Median age in years (IQR / range)
Total	168	92 (55%)	6.3 (3.3 to 10.3 / 0.2 to 37.2)
SCN1A-DEE (Dravet Syndrome)	59	33 (56%)	8.1 (4.8 to 16.3 / 0.4 to 37.2)
KCNB1-DEE	31	17 (55%)	8.2 (4.5 to 12.0 / 2.6 to 22.9)
KCNQ2-DEE	78	42 (54%)	4.5 (2.0 to 8.4 / 0.2 to 20.3)

Most participants (127/168, 76%) were from North America, while 27/168 (16%) were from Europe or the United Kingdom and 14/168 (8%) were from other countries or did not specify. The majority of patients with DEE self-identified White or Caucasian (154/168; 92%). Other reported races were Asian (9/168; 5%); American Indian, Alaskan Native, First Nations (1/168; 0.006%); Black or African American (1/168; 0.006%), and Native Hawaiian or other Pacific Islander (1/168; 0.006%). Twelve (7%) parents reported that their child was of Hispanic or Latin ethnicity.

Gastrointestinal symptoms:

Overall, 92/168 (55%) parents reported constipation or dysmotility in their child. The prevalence of GI symptoms differed significantly across the three DEE types (Table 2). Despite the similar distributions of males and females between the three cohorts (Figure 1A), GI symptoms were not clearly associated with sex (P = 0.08). We explored sex differences within each DEE group and found a substantial association between GI symptoms and sex in the SCN1A-DEE cohort (20/33 [61%] female SCN1A-DEE patients versus 6/26 [23%] male patients [P = 0.004; Figure 1B]). No such association was apparent in the other two DEE groups. Despite age differences across the DEE groups, the age distributions of patients with and without GI symptoms were very similar with median ages of 6.2 and 6.3 years, respectively (Figure 1C). Furthermore, after adjustment for DEE type, patient age was not significantly associated with the manifestation of GI symptoms (P = 0.09, Figure 1D).

Table II.

Distribution of GI-related measures within DEE cohorts^*

	SCN1A-DEE (Dravet Syndrome)	KCNB1-DEE	KCNQ2-DEE	Across-group differences
GI symptom prevalence	(26/59) 44%	(11/31) 35%	(55/78) 71%	P < 0.001
GI symptom frequency % daily % weekly % monthly % <1 day/month	(4/25) 16% (11/25) 44% (5/25) 20% (5/25) 20%	(2/11) 18% (6/11) 55% (2/11) 18% (1/11) 9%	(18/50) 36% (22/50) 44% (9/50) 18% (1/50) 2%	P = 0.14
GI symptom effects
% discomfort	(20/26) 77%	(10/11) 91%	(50/55) 91%	P = 0.09
[a] % Yes, common occurrence/serious concern	(3/26) 12%	(4/11) 36%	(22/55) 40%
[b] % Yes, occasionally but not a serious concern	(17/26) 65%	(6/11) 55%	(28/55) 51%
% appetite disturbances	(12/26) 46%	(8/10) 80%	(32/55) 58%	P = 0.10
[a] % Yes, common occurrence/serious concern	(2/26) 8%	(4/10) 40%	(7/55) 13%
[b] % Yes, occasionally but not a serious concern	(10/26) 38%	(4/10) 40%	(25/55) 45%
Functional mobility % dependent % requires aids % independent	(2/54) 4% (2/54) 4% (50/54) 92%	(6/31) 19% (4/31) 13% (21/31) 68%	(21/50) 42% (7/50) 14% (22/50) 44%	P < 0.001
Ketogenic diet (KD) % currently using KD % not currently using KD	(3/59) 5% (56/59) 95%	(4/31) 13% (27/31) 87%	(2/70) 3% (68/70) 97%	P = 0.13
Gastrostomy tube % exclusive % partial use % never	(2/59) 3% (4/59) 7% (53/59) 90%	(4/31) 13% (3/31) 10% (24/31) 77%	(6/70) 9% (2/70) 3% (62/70) 88%	P = 0.28
Chewing/swallowing difficulties % every meal % daily % weekly % <1 day/week	(8/57) 14% (2/57) 4% (7/57) 12% (40/57) 70%	(0/26) 0% (5/26) 19% (4/26) 15% (17/26) 66%	(14/64) 22% (17/64) 26% (5/64) 8% (28/64) 44%	P = .0018
Autism diagnosis/features	(30/59) 51%	(15/31) 48%	(27/70) 39%	P = 0.35
Seizure prevalence	(56/59) 95%	(23/31) 74%	(26/59) 44%	P < 0.001

^*Due to rounding errors, not all values may add up to 100%. Across-group differences determined using Chi-square tests.

Figure 1.

Sex and age distributions of patients with reported GI symptoms (GI-SX⁺) and those with no reported GI symptoms (GI-SX⁻) within SCN1A-, KCNB1-, or KCNQ2-DEE cohorts. (A) The distribution of females and males did not significantly differ overall between the GI-SX⁺ and GI-SX⁻ patient groups. (B) There were proportionally more females than males with GI symptoms within the SCN1A-DEE cohort, but there were no significant differences between the proportions of males and females with GI symptoms in the KCNB1-DEE or KCNQ2-DEE cohorts. (C, D) The age distribution did not significantly differ between GI-SX⁺ and GI-SX⁻ patient groups in any of the three cohorts. ** = P < 0.01.

Most parents (86/92) who reported GI symptoms in their children also provided information on how frequently these symptoms manifested. Survey responders reported either daily or weekly symptoms in 63/86 (73%) DEE patients overall. Daily or weekly symptoms were reported in 15/25 (60%) SCN1A-DEE patients, 8/11 (73%) KCNB1-DEE patients, and in 40/50 (80%) KCNQ2-DEE patients (Table 2).

Patient discomfort, stemming from the GI symptoms, was considered a common occurrence or serious concern in 29/92 (32%) affected patients (Table 2). In a post-hoc comparison, SCN1A-DEE patients were significantly less likely than KCNB1- and KCNQ2-DEE patients (combined) to experience common or serious discomfort as a result of their GI symptoms (P = 0.02). Parent-reported changes to appetite due to GI symptoms were a common occurrence or serious concern for 13/91 (14%) of responders (Table 2). KCNB1-DEE patients were marginally more likely to experience appetite disturbances as a consequence of GI symptoms, compared with the other two groups (P = 0.06).

Multiple treatments were utilized by parents to relieve the GI symptoms; 40/92 (43%) reported using including dietary supplements, 19/92 (21%) reported using enemas, 44/92 (48%) used medications, and 32/92 (35%) reported using other therapeutics.

Mobility limitations and GI symptoms:

The distribution of functional mobility scores differed significantly across the three cohorts (Table 2). However, using multiple logistic regression to control for the effect of individual DEEs, we found no overall association between functional mobility and presence of GI symptoms (P = 0.47, Figure 2A). Notably, although KCNQ2-DEE patients exhibited both the lowest functional mobility scores (P < 0.001) as well as the highest prevalence of GI symptoms among the three cohorts, within the KCNQ2-DEE group, there was no significant association between mobility scores and GI symptoms (P = 0.23).

Figure 2.

Distributions of SCN1A-, KCNB1-, or KCNQ2-DEE patient functional domains in patient groups with (GI-SX⁺) and without (GI-SX⁻) GI symptoms. Parents reported information regarding the (A) degree of independent mobility, (B) current use of a special diet, (C) relative reliance on a gastronomy tube for feeding, (D) frequency of chewing and/or swallowing difficulties, (E) autism spectrum disorder (ASD) diagnosis or features in their child, and (F) occurrence of any seizures in the previous six months.

Ketogenic diet and GI symptoms:

In total, 9/160 (6%) children were reported to be currently treated with the ketogenic diet (Table 2). On bivariate analysis, use of the ketogenic diet was not associated with GI symptoms (P = 0.14); however, after adjustment for DEE, use of ketogenic diet was significantly associated with GI symptoms (P = 0.03). This change appears to be due to a finding in the KCNB1 cohort; all children in that group who were using the ketogenic diet also had GI symptoms (Figure 2B).

Gastrostomy tube and GI symptoms:

Current partial or exclusive use of a gastrostomy tube was reported in 21/160 (13%) DEE patients and did not differ significantly between DEE groups (Table 2). After adjusting for DEE, use of gastrostomy tube was found to be significantly associated with GI symptoms (P = 0.03; Figure 2C).

Chewing and swallowing difficulties and GI symptoms:

Excluding patients who were exclusively tube-fed, frequent (occurring at least once per meal) difficulties with chewing or swallowing were reported for 22 of 147 (15%) children with DEE (Table 2). The frequency of reported chewing or swallowing difficulties varied significantly across the three DEE groups (Table 2), with KCNB1-DEE patients significantly less likely to have chewing or swallowing difficulties (P = 0.003). After adjusting for DEE, chewing and swallowing difficulties were not associated with reports of GI symptoms (P = 0.52; Figure 2D).

Autism spectrum disorder (ASD) and GI symptoms:

Either a diagnosis or having features of ASD was reported in 72/160 (45%) DEE patients; this did not vary across the three DEE groups (Table 2). After adjusting for DEE, there was no significant association between reported ASD diagnosis or features and GI symptoms (P = 0.08; Figure 2E).

Seizures and GI symptoms:

While seizure prevalence in the previous six months differed significantly between the different DEE groups (Table 2), there was no association between seizures and GI symptoms after adjusting for DEE type (P = 0.99; Figure 2F). Overall, 15/135 (11%) patients were reported to have bowel movements as a seizure trigger, and DEE patients with GI symptoms were significantly more likely to report seizures triggered by bowel movements (P = 0.02).

Anti-seizure drugs and GI symptoms:

Overall, no individual medication appeared to be linked with GI symptoms across the three DEE groups (Figure 3A). In exploratory analyses within each cohort (Figure 3B); we found stiripentol, a drug specifically approved for Dravet syndrome, was significantly associated with increased report of GI symptoms in SCN1A-DEE patients (P < 0.001). We also found a significant association between cannabidiol (CBD) and GI symptoms in KCNB1-DEE patients (P = 0.006) but not the other groups.

Figure 3.

Incidence of GI symptoms in SCN1A-, KCNB1-, or KCNQ2-DEE patients taking anti-seizure medications. Relative risk (RR) with 95% CI was calculated for (A) medications being taken by ≥ 10 patients across all three disease cohorts as well as (B) medications being taken by ≥ five patients within each respective DEE cohort. Higher RR values are associated with an increased incidence of GI symptoms in patients currently taking a particular medication. ** = P < 0.01, *** = P < 0.001.

Discussion

Children with ion channel-gene-linked DEEs have a high prevalence of GI symptoms, specifically constipation and dysmotility, which do not appear to be explained by common risk factors that frequently arise in the DEE population, such as limited functional mobility, anti-seizure medications, or presence of ASD. These symptoms create additional needs for medical care, place burdens on parents and caregivers, and negatively impact the quality of the patients’ lives.

By initially focusing on DEE patients with variants in the genes encoding a subset of Na_v and K_v channels, we have documented the prevalence and risk factors for GI symptoms in channelopathy-linked DEE patients. Although we have no direct comparison to the frequency of similarly reported GI symptoms in a healthy population, a meta-analysis of 18 independent studies reporting on constipation in the general population found a median prevalence of constipation in children of ~9% (interquartile range: 5% to 17%), with the highest estimate being 30% (40). These estimates from population-based studies are well below our overall finding of 55% (95% confidence interval 47% to 62%) prevalence in DEE patients. Our findings emphasize the prevalence and severity of GI symptoms in channelopathy-associated DEE patients over a broad range of ages, while highlighting potential differences between three different channelopathy cohorts.

While this study is focused on channelopathies, it is important to note that GI symptoms are prevalent in multiple DEEs. We further considered reports of GI symptoms from two non-channelopathy, gene-specific cohorts who had participated in a similar survey with the same questions regarding GI symptoms (in review). Parents of 71 CHD2- and PACS1-linked DEE patients reported rates of constipation and gastric dysmotility in their children, and the average prevalence of these symptoms across these two groups was 39%, which was significantly lower than the average prevalence of GI symptoms, 55%, reported for this study’s channelopathy groups). While relatively little is known about the role of the CHD2 and PACS1 genes in GI function, it is interesting to note that CHD2 variants have been identified in multiple human gastric and colorectal cancers (41). Future research could determine if the GI symptoms in the CHD2- and PACS1-linked DEEs are similarly not explained by common risk factors often exhibited by DEE patients.

Risk factors for GI symptoms are common in the DEE population and include mobility limitations, ketogenic diet, and medications. None of these, however, appeared to be a major driving factor behind the prevalent GI symptoms in our DEE cohorts. For example, the KCNQ2-DEE patients were the least mobile group and exhibited the highest proportion of GI symptoms. Yet, within the KCNQ2-DEE cohort, there was no association between degree of mobility and GI symptoms. Further, there was no overall association between GI symptoms and mobility after adjusting for DEE type. Small, borderline associations were observed for well-known risk factors such as gastrostomy tube usage and ketogenic diet therapy. As so few children had these interventions, their presence does not explain the high reported prevalence of GI symptoms in our study patients. Interestingly, sex was significantly associated with reported GI symptoms in SCN1A-, but not KCNQ2- or KCNB1-DEE patients; females with SCN1A-linked DEE were about three times as likely as their male counterparts to have these symptoms. Though females are more likely than males to develop chronic constipation in the general population (42), this differential was not apparent in the two other DEE groups. Independent replication of the finding is needed, but it could reflect sex-specific aspects of channelopathy pathophysiology.

Although previous research suggests a connection between bowel movements and seizure activity (24), we observed no statistically significant association between GI symptoms and recent seizure activity in these channelopathy cohorts. In a separate part of the survey in which seizure triggers were queried, only a small percentage of parents reported bowel movements as a seizure trigger. Future studies could look more broadly at potential correlations between these symptoms in multiple forms of pediatric epilepsies.

Multiple anti-seizure drugs are used in patients with DEEs, and these medications are known to have a wide range of side-effects, including GI dysfunction. While we are very cautious not to overinterpret our findings due to the small numbers, reporting use of any one medication and the multiple hypotheses tested, a few findings may inform new hypotheses to test formally in future studies. Stiripentol, a drug approved exclusively for Dravet syndrome, was significantly associated with GI symptoms in the SCN1A-DEE group. GI symptoms were also associated with cannabidiol (CBD) use in KCNB1-DEE patients only. Unlike stiripentol, CBD was taken by multiple children from all three DEE cohorts. Whether this particular finding for CBD reflects a specific effect on GI function in the context of K_v2.1 channel impairment remains to be determined.

The prevalence of GI symptomology in DEEs linked to variants in SCN1A, KCNB1, or KCNQ2 is not well explained by common risk factors for GI dysfunction and raises the question of whether the proteins encoded by these genes may contribute to proper GI tract function. SCN1A gene expression has been localized to the human colonic myenteric plexus, the branch of the enteric nervous system that controls bowel motility (27). A study in mice found that mechanosensitive colonic fibers express Na_v1.1 (43). While gastrointestinal expression of KCNB1 and KCNQ2 in humans is not well-characterized, K_v2.1 was found to be localized to murine colonic myenteric plexus (44), and K_v7 channels have been localized to intrinsic neuronal populations as well as smooth muscle in the human GI tract (45). Functionally, K_v7 channels are thought to be important for controlling contractile activity in the murine intestinal tract (38) as well as mediating sensory colonic afferent responses to noxious stimuli in mice and humans (45). Importantly, a wide variety of voltage-gated ion channel-encoding genes that are known to be expressed in the GI tract, in addition to brain, have been linked to various diseases, including other DEEs, that manifest with GI symptomology (Table 3). GI symptoms in channelopathy-associated DEEs are common, serious, but poorly understood comorbidities of these DEEs. The next steps will entail investigating the mechanisms and elucidating the roles of GI-expressed ion channels in GI function and dysfunction. The combination of our survey results and published data, suggest DEE-linked voltage-gated ion channel gene variants may also contribute to the GI dysmotility and constipation commonly reported in patients and lay the groundwork for future mechanistic studies in animal models of DEE.

Table III.

GI symptoms in other channelopathies

Ion channel gene (protein product)	Disease(s) associated with genetic variant that manifest with GI symptoms	Known localization of gene and/or protein product in human GI tract	Known GI symptoms associated with genetic variant
SCN1B (Nav β1/β1 subunits)	OMIM EIEE52 (46)	Colonic ICCs, colonic PDGFRα+ cells, colonic MP, colonic SP, colonic smooth muscle (47)	Constipation, FTT (A. Poduri, pers. comm., 6 May 2020)
SCN2A (Nav1.2)	OMIM EIEE11 (2, 8)	Colonic MP, colonic smooth muscle (27)	Constipation, diarrhea, FTT, vomiting (2, 8)
SCN3A (Nav1.3)	OMIM EIEE62 (48)	Colonic MP, colonic smooth muscle, enterochromaffin cells in small intestine and colon (27, 49)	FTT (48)
SCN4A (Nav1.4)	NDM (6)	Colonic smooth muscle (27)	Constipation, FTT, gastroesophageal reflux (6)
SCN5A (Nav1.5)	IBS, LQTS (33, 50-52)	Colonic MP, ICCs in small intestine, smooth muscle in small intestine and colon (27, 28, 30, 32, 34, 35)	Abdominal pain, constipation, diarrhea (33, 50-52)
SCN8A (Nav1.6)	OMIM EIEE13 (13)	Colonic MP (27)	Constipation, gastroesophageal reflux (13)
SCN9A (Nav1.7)	NPS/SFN (3, 7, 25)	Colonic MP, colonic smooth muscle (27)	Abdominal/rectal pain, constipation, diarrhea, (3, 7, 25)
SCN10A (Nav1.8)	NPS/SFN (4, 25)	Colonic MP (27)	Abdominal discomfort, constipation, diarrhea, gastroesophageal reflux, gastroparesis, vomiting (4, 25)
SCN11A (Nav1.9)	CIP, NPS/SFN (9, 25, 53)	Colonic MP, colonic SP, colonic smooth muscle (27, 31)	Abdominal/perineal/rectal pain, constipation, diarrhea, enlarged colon, FTT, intestinal dysmotility (9, 25, 53)

Abbreviations: CIP = congenital insensitivity to pain; EIEE = early infantile epileptic encephalopathy; FTT = failure-to-thrive; IBS = irritable bowel syndrome; ICCs = interstitial cells of Cajal; LQTS = congenital long QT syndrome; MP = myenteric plexus; NDM = non-dystrophic myotonia; NPS = neuropathic pain syndromes; PDGFRα⁺ = platelet-derived growth factor receptor α⁺; SFN = small fiber neuropathy; SP = submucosal plexus

Several limitations of our study must be acknowledged. First, the data presented are derived from parent reports. This includes the diagnosis of specific ion channel variants linked to each child’s DEE, although several parents provided genetic testing reports that corroborated the etiology. Because the families were recruited through groups that were specific to their child’s disease, we expect any error in reporting for etiology to be minimal. In addition, the exact description of GI symptoms alone may not have been sufficiently specific and may have reflected how such questions are asked in a simple review of systems often performed in a medical office. Additional questions regarding frequency, discomfort, and impact on appetite helped to define the scope of and contextualize the problem in the children’s and parents’ daily lives. Information about potential risk factors for GI symptoms was also obtained by report; however, the information was appropriate to be obtained in this manner. For functional mobility, we utilized the GMFCS, which has been previously validated for parent report (39). Other questions were developed in collaboration with parents and based on review of similar measures available throughout the literature. Finally, some questions were straightforward checklist items, such as whether a child had a gastrostomy tube or was being treated with the ketogenic diet. We also acknowledge that this survey was disseminated to an unknown number of people. We therefore cannot be certain that the responses of our 168 parent respondents can be extrapolated to the general population of children with DEEs. Notably, this is often the case in studies of rare diseases. The use of a web-based survey, however, did permit us to reach a large number of people, much greater than would have been possible in a single or even multi-center study over a relatively brief period of time. Some of our comparisons, particularly medications, were exploratory. We had the opportunity to ask these questions but often did not have the statistical power to address them. Further, the number of medications reported would reasonably raise concerns regarding multiple hypothesis testing. Thus, these results should be considered hypothesis-generating only. In this survey, we directly asked only about current use of medications for treatment of seizures and GI symptoms. We therefore do not have information on other medications with potential GI side-effects that may be used by children in this cohort for treatment of other DEE comorbidities.

In conclusion, this work provides novel information that highlights the importance of GI dysfunction in the DEEs and suggests these GI symptoms may be part of the disease itself and not just a result of poor mobility, medications, or other factors. Elucidating the role of ion channel dysfunction in GI symptoms for patients with DEE may guide future therapeutic efforts to alleviate this common and often distressing symptom.

Recurring Abbreviations and Acronyms:

DEE

Developmental and epileptic encephalopathy

GI

Gastrointestinal

ASD

autism spectrum disorder

Data Accessibility Statement:

Due to privacy and ethical concerns, limited supporting data may be made available only to bona fide researchers subject to an appropriate data use agreement. Details of the data set and how to request access are available from Dr. Anne Berg at Ann & Robert H. Lurie Children’s Hospital of Chicago.