MED13L-related disorder characterized by severe motor speech impairment

Abstract

Background

MED13L-related disorder is associated with intellectual disability, motor delay, and speech deficits. Previous studies have focused on broad clinical descriptions of individuals, but limited information regarding specific speech diagnoses and results of direct testing has been published to date. We conducted deep phenotyping to characterize the speech, language, motor, cognitive, and adaptive phenotypes of individuals with MED13L-related disorder.

Methods

In this cross-sectional study, we administered standardized articulation, language, motor, and cognitive testing to 17 children and adolescents (mean age 9y 9m; SD 4y 5m; range 4y 2m to 19y 7m). In-person testing was supplemented with broad developmental, medical, and behavioral information collected virtually from a cohort of 67 individuals.

Results

All individuals who completed in-person articulation testing met diagnostic criteria for speech apraxia, dysarthria, or both. Language impairment was present in all of the in-person cohort and reported for almost all (97%) of the virtual cohort. Those who were able to complete motor testing demonstrated significant deficits in visual motor integration (mean 57.08, SD 9.26). Full scale IQs fell in the borderline to intellectual disability range, consistent with reported cognitive impairment in 97% of the virtual cohort. Notable medical features included hypotonia (83%), vision problems (72%), recurrent otitis media (58%), gastrointestinal problems (57%), and seizures (31%).

Conclusions

MED13L-related disorder is characterized by a high rate of motor speech disorders that occur in the context of globally impaired motor, language, and cognitive skills. Children would benefit from early referrals to speech therapy to assess their speech, language, and support needs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s11689-025-09645-1.

Background

The MED13L (Mediator Complex Subunit13-like) gene encodes one component of the Mediator transcription co-activator complex, which serves as a bridge between RNA polymerase II and transcription factors [1, 2]. It is highly expressed in the fetal and adult brain (particularly the cerebellum), skeletal muscle, and heart [3]. Knockdown of MED13L deregulates genes within the Wnt signaling pathway, among others, and leads to the impaired development of neural crest cells [1, 2].

Initially, rare heterozygous missense variants in MED13L were identified in several patients with cardiac defects (dextra-looped transposition of the great arteries) [3]. A later study also identified conotruncal heart defects and intellectual disability in patients with haploinsufficiency of MED13L [4]. Subsequently, more than 100 patients with disease-causing variants in the MED13L gene have been described. Loss of function variants, including whole gene or intragenic deletions, are most common, although there have been recent reports of missense variants associated with higher rates of epilepsy and a more severe phenotype [5, 6]. Pathogenic variants almost always occur de novo, although there have been three case reports of germline mosaicism [7–9]. MED13L-related disorder is rare, occurring in approximately 6 in 100,000 live births by one estimate [10].

Despite the initial reports of heart defects, cardiac anomalies have been reported in a minority of patients with MED13L pathogenic variants to date. The most consistently reported features associated with MED13L-related disorder are intellectual disability, motor delay, and speech impairment. Additional commonly reported features include facial dysmorphism (macroglossia, macrostomia, depressed nasal root, ear anomalies), hypotonia, and seizure disorder [5, 6, 11]. Although severe speech impairment and motor deficits are frequently reported, few specific details regarding the speech and motor phenotype have been provided. The two largest published case series both report severely impaired or absent speech and motor delay in most patients, but do not include results of direct testing or specific speech, language, or motor diagnoses in the majority of individuals [6, 11]. Most case reports do not differentiate between “speech” and “language” delay, although this is an important distinction that is worth noting because speech and language domains can be differentially impacted. Language refers to the abstract, rule-based system of meaning, structure, and social communicative properties of words, whereas speech is the spoken modality of language, requiring physical movement and coordination of the articulators, such as the lips and tongue. As such, there is a motor component to speech production that should be considered.

Speech disorders in children can be caused by many factors, but the prevalence and co-occurrence of both speech and motor impairment in MED13L-related disorder raises suspicion for possible pediatric motor speech disorders (MSDs), which are rare but intractable speech disorders of childhood [12]. The two main pediatric MSDs are childhood apraxia of speech (CAS) and childhood dysarthria. CAS is caused by disrupted motor planning and programming [13], whereas childhood dysarthria arises from deficits in motor execution resulting from impaired muscle strength, tone, or coordination [14]. Indeed one patient with MED13L-related disorder was reported to have a diagnosis of “verbal dyspraxia” [11], an alternative name for CAS, and another case series described one pediatric patient with MED13L-related disorder who had been diagnosed with dysarthria [6]. Additionally, a recent study examining the results of exome sequencing in a large cohort of children with MSDs identified causative MED13L variants in 6 children with CAS [15]. However, no detailed phenotypic information or standardized test results were available for these patients beyond general clinical description.

Granular information about speech production in individuals with MED13L-related disorder is missing from the literature, but would be useful for genetic counseling and treatment planning with this population. Deep speech phenotyping studies have broader utility as well because they can drive hypotheses regarding the relationships between genes, brain expression, and the pathophysiology of speech and language disorders [16]. This study seeks to bridge this gap in the literature by collecting detailed information about the motor speech phenotype of individuals with MED13L-related disorder. Possible MSDs are considered in the context of language, motor, cognitive, adaptive, and behavioral features of individuals with MED13L-related disorder, collected through detailed articulation, language, psychological, and motor assessments. Deep phenotyping was conducted through in-person testing for a subset of individuals, supplemented with broad developmental, medical, and behavioral information from a larger group ascertained from Simons Searchlight [17], a large online cohort of individuals with genetic changes that impact development.

Methods

Patients

In-person cohort

Sixteen individuals were recruited through the 2022 Simons Searchlight Family Meeting, and one additional individual was recruited through an outpatient pediatric neurodevelopmental clinic. Eligibility criteria included age (between 2 and 21 years) and confirmed pathogenic or likely pathogenic (P/LP) MED13L variant, interpreted using the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) criteria [18]. The study was conducted under a minimal risk protocol as defined by the Common Rule, and written, informed consent was obtained for all patients or their legal representatives (45 CFR 46.116).

Simons searchlight cohort

Sixty-seven individuals with P/LP MED13L variants joined the Simons Searchlight study and completed at least some surveys, interviews, and questionnaires. The number of individuals who completed each virtual measure through the Simons Searchlight study varied, as not everybody completed all surveys or questionnaires that were available to them. Genetic test reports (exome sequencing or gene panel sequencing) were reviewed for eligibility using ACMG/AMP criteria [18]. No individuals with variants of uncertain significance were included. Study data were obtained from a public data clearinghouse, SFARI base (base.sfari.org), with proof of related IRB approval. Thirteen individuals who participated in the in-person protocol also participated in the Simons Searchlight virtual cohort (Table 1).

Table 1.

MED13L variants of the in-person cohort

ID	Variant (NM_015335.4)	Type	Inheritance	Classification	Age at Diagnosis
F1S1a	c.5990dup,p.Leu1997Phefs*19	LoF	De novo	Pathogenic	3y8m
F2S1	c.401dup,p.Met135Asnfs*3	LoF	De novo	Pathogenic	2y5m
F3S1a	c.5471del,p.Asn1824Metfs*28	LoF	Inherited—Paternal (somatic mosaic; 7% of reads)	Pathogenic	3y8m
F4S1a	c.2597C > T,p.Pro866Leu	Missense	De novo	Pathogenic	0y7m
F5S1a	arr[GRCh37] 12q24.21(116,519,929–116,553,225) × 1	Deletion	Unknown	Pathogenic	17y6m
F6S1a	c.6556C > T,p.Gln2186*	LoF	De novo	Likely Pathogenic	2y11m
F7S1a	c.3469C > T,p.Gln1157*	LoF	De novo	Pathogenic	2y9m
F8S1a	arr[GRCh37] 12q24.21(116,673,140–116,676,995) × 1	Deletion	De novo (germline mosaicism inferred)	Pathogenic	4y10m
F8S2a	arr[GRCh37] 12q24.21(116,673,140–116,676,995) × 1	Deletion	De novo (germline mosaicism inferred)	Pathogenic	6y6m
F9S1	c.4077G > A,p.Trp1359*	LoF	De novo	Pathogenic	5y8m
F10S1	arr[GRCh37] 12q24.21(116476163_116619234) × 1	Deletion	De novo	Pathogenic	1y6m
F11S1a	arr[hg18] 12q24.21(114,895,862–115,110,038) × 1	Deletion	De novo	Pathogenic	3y0m
F12S1a	c.4609C > T,p.Gln1537*	LoF	De novo	Pathogenic	1y5m
F13S1a	c.4403dup,p.Thr1470Asnfs*9	LoF	De novo	Pathogenic	6y7m
F14S1a	c.1690C > T,p.Arg564*	LoF	De novo (germline mosaicism inferred)	Pathogenic	5y0m
F14S2a	c.1690C > T,p.Arg564*	LoF	De novo (germline mosaicism inferred)	Pathogenic	1y10m
F15S1	c.6488C > T,p.Ser2163Leu	Missense	De novo	Likely Pathogenic	13y2m

Abbreviations: LoF loss of function, y years, m months

^aIndividuals also completed some aspects of Simons Searchlight virtual assessment protocol

In-person assessment protocol

Speech

Fourteen individuals produced enough verbal speech to complete the speech assessment, which was administered by a certified speech-language pathologist with expertise in pediatric MSDs. Three children did not participate in articulation testing because they used fewer than 10 verbal words and were not able to consistently attempt to repeat words on request.

The Sounds in Words subtest of the Goldman-Fristoe Test of Articulation- Third Edition (GFTA-3) [19] was administered to all 14 individuals and completed in its entirety by 12 children (two additional children could not complete the test due to fatigue). The GFTA-3 assesses a child’s production of phonemes across initial, medial, and final positions of words. It provides norm-referenced standard scores (mean 100, S.D. 15) and growth scale values for individuals aged 2–21 years. The GFTA-3 Sounds in Words subtest was selected due to the simple instructions/low task demands, wide normative age range, and ability to calculate both standard scores and growth scale values. The Dynamic Evaluation of Motor Speech Skill (DEMSS) [20] was administered to a subset of four children to gather additional information regarding core features of CAS, including inconsistency, disrupted prosody, and vowel errors. The DEMSS is a criterion-referenced measure for young and/or severely speech-impaired children in which lower scores are associated with an increased likelihood that a diagnosis of CAS is correct. Scores are broadly separated into three groups: 426–373 (Little or no evidence of CAS), 372–323 (Some evidence for at least mild CAS), and 322–0 (Significant evidence for CAS). A naturalistic speech sample was transcribed for all 14 individuals and analyzed for features of CAS and childhood dysarthria. The Profile of Childhood Apraxia of speech and Dysarthria (ProCAD) [21] was utilized to aid with differential diagnosis of individuals’ speech disorders. The ProCAD is a diagnostic protocol that provides clinicians with a speech feature checklist and decision tree to aid in the differential diagnosis of CAS vs pediatric dysarthria.

A diagnosis of CAS was made based on the presence of all three consensus criteria from the American Speech-Language-Hearing Association, including inconsistent errors, lengthened and disrupted coarticulatory transitions, and inappropriate prosody [13]. Additionally, the child also needed to demonstrate features associated with CAS only in both the articulatory and rate/prosody domains, as outlined in the ProCAD. A diagnosis of childhood dysarthria was assigned based on imprecise articulatory contacts or the presence of multiple dysarthria-only features across speech subsystems, following the decision tree provided in the ProCAD.

Language

Two standardized vocabulary tests, the Peabody Picture Vocabulary Test, Fifth Edition (PPVT-5) [22] and Expressive Vocabulary Test, Third Edition (EVT-3) [23] were administered by a certified speech-language pathologist to evaluate receptive and expressive vocabulary skills, respectively. The standard scores (mean 100, S.D. 15) yielded from these single-word vocabulary tests were derived from the same normative sample, allowing direct comparisons between individuals’ receptive and expressive vocabulary skills. Although limited in scope, the PPVT-5 and EVT-3 were utilized due to their short administration time, simple instructions, wide age range (normed for 2 years, 6 months to 90 + years), and ability to directly compare receptive and expressive skills. Individuals were determined to have a language disorder if they scored below one S.D. on either measure.

Cognition

A pediatric psychologist or trained research assistant administered one of three cognitive tests to all individuals, depending on age and ability level: the Differential Abilities Scale, Second Edition (DAS-2) [24] Kaufman Brief Intelligence Test, Second Edition (KBIT-2) [25]; or the Mullen Scales of Early Learning (MSEL) [26]. All tests administered yield standard scores; however, some individuals in this study were administered an assessment outside of their range due to significant impairments in cognitive and/or verbal abilities. In these cases, a developmental quotient using age equivalent scores was used to approximate IQ.

Visual motor

The Beery-Buktenica Developmental Test of Visual Motor Integration, Sixth Edition (Beery VMI) [27] was administered by a pediatric psychologist or trained research assistant to all individuals and was completed in its entirety by 12 children. The Beery VMI is appropriate for use in individuals ages 2 and older and measures visual motor skills through a drawing activity. The Beery VMI measures visual motor integration and is a commonly used measure of fine motor skills.

Virtual assessment protocol

Medical and developmental

Simons Searchlight genetic counselors administered standardized, qualitative medical and developmental histories by telephone interview with primary caregivers of individuals with MED13L-related disorder. A seizure history survey was obtained by online questionnaire. One individual was recruited from a neurodevelopmental clinic and medical history data was ascertained through review of electronic health record (EHR) data.

Adaptive

The Vineland Adaptive Behavior Scales, Third Edition (VABS) [28] was completed by online questionnaire by primary caregivers. The VABS Parent/Caregiver form assesses a person’s adaptive skills that are important for functional daily living. Standard scores (mean 100, S.D. 15) were calculated across four adaptive domains: Communication, Daily Living Skills, Socialization, and Motor Skills, the first three of which were used to calculate a person’s overall Adaptive Behavior Composite (ABC) score.

Behavior

The Child Behavior Checklist (CBCL) [29, 30] was completed by online questionnaire by primary caregivers. Two forms of the CBCL were administered based on the age of the child (ages 1.5–5 and ages 6–18). The CBCL is a standardized questionnaire that assesses a child’s behavioral problems across eight syndrome scales (Aggressive Behavior, Anxious/Depressed, Attention Problems, Rule-Breaking Behavior, Somatic Complains, Social Problems, Thought Problems, Withdrawn/Depressed) and five to six DSM-5-Oriented Scales (Depressive Problems, Anxiety Problems, Autism Spectrum Problems [ages 1.5–5], Somatic Problems [ages 6–18] Attention Deficit/Hyperactivity Problems, Oppositional Defiant Problems, Conduct Problems [ages 6–18]). These subdomains are used to calculate broad scores for Internalizing Behaviors, Externalizing Behaviors, and Total Problems. The CBCL reports T-scores in which < 65 is interpreted as ‘Normal,’ 65–69 is interpreted as ‘Borderline’ clinical range, and ≥ 70 is considered the ‘Clinical’ range.

Results

In-person cohort

Seventeen individuals (10 female, 7 male) from 15 families participated in the in-person assessment, including two sets of siblings (Table 2). The mean age was 9 years, 9 months (range 4y 2 m to 19y 7 m; SD 4y 5 m). Sixteen of 17 variants were de novo, with inferred germline mosaicism for the two families with affected siblings (Table 1). The single case of an inherited variant was from a father with somatic mosaicism (pathogenic variant identified in 7% of reads). There were 12 sequence variants (10 nonsense/frameshift, 2 missense) and 5 intragenic deletions (Fig. 1). The mean age of genetic diagnosis was 4 years, 11 months (range 0y 7 m – 17y, 6 m).

Table 2.

Assessment protocol for in-person and virtual cohorts

Cohort (N)	Domain	Assessment Protocol (n)
In-Person (17)	Speech	GFTA-3 (12)
		DEMSS (4)
		ProCAD (14)
	Language	PPVT-5 (14)
		EVT-3 (13)
	Cognition	MSEL (3)
		DAS-2 (6)
		KBIT-2 (6)
	Visual Motor	Beery VMI (15)
Virtual (67)	Medical and Developmental	Medical history interview (59)
		Seizure survey (65)
		Developmental history interview (67)
		EHR review (1)
	Adaptive	VABS-3 (47)
	Behavior	CBCL 1.5–5 years (31)
		CBCL 6–18 years (25)

Abbreviations: GFTA-3 Goldman Fristoe Test of Articulation, Third Edition, DEMSS Dynamic Evaluation of Motor Speech Skill, ProCAD Profile of Childhood Apraxia of Speech and Dysarthria, PPVT-5 Peabody Picture Vocabulary Test, Fifth Edition, EVT-3 Expressive Vocabulary Test, Third Edition, MSEL Mullen Scales of Early Learning, DAS-2 Differential Abilities Scale, Second Edition, KBIT-2 Kaufman Brief Intelligence Test, Second Edition, Beery VMI Beery-Buktenica Developmental Test of Visual Motor Integration, EHR Electronic health record, VABS-3 Vineland Adaptive Behavior Scale, Third Edition, CBCL Child Behavior Checklist

Fig. 1.

MED13L (NM_015335.5) exons, functional domains, and pathogenic variants of in-person cohort were plotted using the trackViewer library [31]. Note that intragenic deletions are shown relative to exons although they also span introns, thus their sizes are not to scale in this figure

Speech

Three children did not produce enough verbal speech to complete articulation testing; that is, they produced fewer than 10 verbal words and were unable to consistently repeat words on request. The remaining 14 who participated in articulation testing met diagnostic criteria for at least one MSD: 7 with CAS only, 6 with CAS + childhood dysarthria, and 1 with childhood dysarthria only (Table 3). On average, individuals exhibited 10 characteristic features of CAS or childhood dysarthria recorded by the ProCAD (range 5–15; SD 2.88) (Table S1, Supplementary Material). Twelve participants completed the GFTA-3, and all individuals’ raw scores corresponded to the lowest achievable standard score (40). Because of this floor effect, growth scale values rather than standard scores are reported in Table 3. Two additional children completed a portion of the GFTA-3 but were unable to complete it in its entirety due to fatigue, so standard scores could not be calculated for them. Four participants also completed the DEMSS, in which lower scores are associated with higher likelihood of a CAS diagnosis. Participants’ raw scores on this measure ranged from 358 (evidence for at least mild CAS) to 114 (significant evidence for CAS). Seven children used some form of augmentative and alternative communication (AAC), including speech-generating devices (6 children) and sign language (1 child). An additional child had attempted to use AAC but had not been successful. While non-speech oral-motor skills were not measured comprehensively, families reported a high rate of dysfunction in this area, including difficulty feeding in infancy (6 children), drooling (4 children), and open mouth posture/tongue thrust (3 children).

Table 3.

Speech and oral-motor phenotype

ID	CAS	Dysarthria	GFTA-3 growth scale value	DEMSS raw score	Oral-motor dysfunction	Verbal Speech	AAC
F1S1	+	-	452	NA	-	Single words, some phrases	-
F2S1	+	-	NA	114	-	Single words	Uses SGD
F3S1	+	+	460	NA	Drooling	Single words	-
F4S1	NA	NA	NA	NA	Drooling; open mouth posture; difficulty feeding in infancy	Minimally verbala	Uses SGD
F5S1	+	+	507	NA	-	Conversational	-
F6S1	+	-	472	NA	Oral hypotonia; open-mouth posture	Single words	-
F7S1	NA	NA	NA	NA	-	Minimally verbala	Attempted but not successful
F8S1	+	-	441	NA	-	Single words	Uses SGD
F8S2	NA	NA	NA	NA	Difficulty feeding in infancy	Minimally verbala	Uses SGD
F9S1	-	+	465	NA	-	Simple phrases, sentences	Uses sign language
F10S1	+	-	NA	NA	Difficulty feeding in infancy	Single words	-
F11S1	+	+	515	356	-	Conversational	-
F12S1	+	+	497	NA	Drooling; tongue thrust; difficulty feeding in infancy	Single words	-
F13S1	+	+	515	353	Difficulty feeding in infancy	Simple phrases, sentences	-
F14S1	+	-	491	NA	Drooling	Single words	Uses SGD
F14S2	+	-	481	358	-	Simple phrases, sentences	-
F15S1	+	+	441	NA	Difficulty feeding in infancy, history of dysphagia	Single words, some phrases	Uses SGD

Abbreviations: CAS childhood apraxia of speech, GFTA-3 Goldman Fristoe Test of Articulation, 5th Edition, DEMSS Dynamic Evaluation of Motor Speech Skills, AAC augmentative and alternative communication, SGD speech-generating device

^aUses < 10 verbal words and unable to consistently repeat words on request

Language

Fourteen participants completed the PPVT-5, and 13 participants completed the EVT-3. All participants scored below the average range on both receptive and expressive vocabulary tests (Table 4). The average standard score on the PPVT-5 was 54.86 (range 40–79; SD 12.04) and the average standard score on the EVT-3 was 60.00 (range 40–80; SD 11.92). Most participants (8 of 13) who completed both vocabulary tests did not exhibit a significant difference between their scores on the PPVT-5 and EVT-3. Out of 17 individuals who participated in at least some aspects of the in-person assessment, 3 children were minimally verbal (fewer than 10 words), 12 children exhibited single-word or phrase-level speech, and 2 children were conversational.

Table 4.

Cognitive, language, visual-motor testing standard scores

ID	Age in months	FSIQ	Verbal IQ	Nonverbal IQ	PPVT-5	EVT-3	Beery VMI
F1S1	97	U	U	U	62	58	U
F2S1	73	64	43	85	56	50	65
F3S1	58	67	57	77	62	65	56
F4S1	67	49	20	20	50	U	U
F5S1	213	40	42	44	40	50	45
F6S1	94	48	46	69	40	67	58
F7S1	88	10a	U	U	U	U	U
F8S1	134	20a	U	U	40	40	U
F8S2	154	8a	U	U	51	67	45
F9S1	116	35a	U	U	U	U	56
F10S1	50	54	45	66	U	U	72
F11S1	155	54	63	56	59	67	57
F12S1	67	M	M	M	79	80	M
F13S1	160	68	65	79	58	69	63
F14S1	138	52	60	55	62	60	70
F14S2	99	72	72	80	69	67	53
F15S1	235	52	50	62	40	40	45
Mean	117.53	46.20	51.18	63.00	54.86	60.00	57.08
SD	53.34	20.20	14.37	18.86	12.04	11.92	9.27

Abbreviations: FSIQ full scale IQ, PPVT-5 Peabody Picture Vocabulary Test, 5th Edition, EVT-3 Expressive Vocabulary Test, 3rd Edition, Beery VMI Beery-Buktenica Test of Visual-Motor Integration, U patient unable to complete, M missing data/test not administered

^aIQ extrapolated from age equivalents on developmental assessment subtests

Cognition

Fifteen individuals completed a cognitive assessment. Due to significant cognitive and verbal impairment, some participants completed a cognitive/developmental assessment that was outside of the normed age range for the test. Three individuals completed the MSEL (mean age 103 months, range 67–155 months), six individuals completed the DAS-2 (mean age 90 months, range 50–134 months), and six individuals completed the KBIT-2 (mean age 144 months, range 99–235 months). Overall, all participants had measured full scale IQs within the range of borderline to intellectual disability (Table 4). The average full scale IQ score was 46.20 (SD 20.20). In cases where both verbal and nonverbal IQ scores were calculable, nonverbal scores (mean 63.00, SD 18.96) were higher than verbal scores (mean 51.18, SD 14.37).

Visual motor

Twelve individuals with MED13L completed the Beery VMI. Three additional individuals attempted but were unable to complete any items on the VMI (would not grasp a pencil or make marks on paper). Of those who completed the Beery VMI, three scored at the floor of the assessment. Overall, all individuals demonstrated significant deficits in visual motor integration (mean 57.08, SD 9.26) (Table 4).

Virtual cohort

Medical and developmental

Caregivers of 59 individuals with P/LP pathogenic variants from Simons Searchlight completed the medical history interview, and one additional individual’s medical information was ascertained through EHR review (N = 60; Table S2 in Supplementary Material). Most individuals (58, 97%) were reported to have at least one neurological feature other than seizures, with low muscle tone being the most common (50, 83%). Additional neurological features included clumsiness/coordination disorder (13, 22%), high muscle tone (8, 13%), movement disorder (6, 10%), large head size (6, 10%), small head size (3, 5%), and cerebral palsy (2, 3%). Gastrointestinal problems, including constipation, gastroesophageal reflux, and diarrhea, occurred in over half of patients (34, 57%). Other prevalent medical features included vision problems (43, 72%) and recurrent otitis media (35, 58%). Cardiac anomalies (including patent ductus arteriosis, tetralogy of Fallot, and bicuspid aortic valve, among others) varied widely and occurred in a minority of individuals (11, 18%).

Sixty-five individuals’ caregivers completed a separate seizure questionnaire, and 67 completed a developmental history interview (Fig. 2A). Seizures were reported in a sizable minority of patients (20/65, 31%), seven of whom required emergency intervention. All but two individuals (97%) with P/LP MED13L variants were reported to have some degree of cognitive impairment, including intellectual disability, developmental delay, and/or learning disability. A similar percentage of individuals also reported a language disorder diagnosis (63/65, 97%), with about a third of children older than 48 months described by their caregivers as minimally verbal (15/49, 31%). Other neurodevelopmental disorders were reported in fewer than half of patients, including autism spectrum disorder (28/57, 49%) and attention deficit/hyperactivity disorder (13/60, 22%).

Fig. 2.

Neurodevelopmental diagnoses and adaptive skills of virtual cohort. A Radar plot displaying the most common neurodevelopmental diagnoses reported by caregivers of children with MED13L pathogenic variants (N = 67). Developmental histories obtained by standardized telephone interview with trained genetic counselors through Simons Searchlight. B Box and whiskers plot displaying the average standard scores of the Vineland Adaptive Behavior Scale for the four main subdomains (Communication, Daily Living Skills, Socialization, and Motor Skills) and the composite score (N = 47). Mean (X), median, interquartile range, and outlier points are depicted. Reference lines designate the normative mean standard score (100), one S.D. below the mean (85) and two S.D.s below the mean (70). Abbreviations: S.D. = standard deviation

Adaptive

Caregivers completed the VABS-3 on 47 individuals with pathogenic mutations in MED13L. Mean standard scores were consistent with significant delays (two or more standard deviations below the normative mean) across all adaptive subdomains (Communication, Daily Living Skills, Socialization, Motor Skills) with no areas of relative adaptive strengths or weaknesses (Fig. 2B; Table S3 in Supplementary Material).

Behavior

Parent report of behavioral symptoms in the 1.5–5-year age range using the CBCL was completed for 31 individuals (Table S4 in Supplementary Material). Results show no mean scores within the clinically elevated range (> 98th percentile) of any scale or subscale (Total Problems mean T-score 58.68, SD 12.93). Mean subclinical elevations (95th-98th percentile) were identified in the domains of withdrawal and attention problems. Parent report of behavioral symptoms in the 6–18 year age range using the CBCL was completed on 25 individuals and no mean scores within the clinically elevated range are present (Total Problems mean T-score 58.50, SD 9.51; Table S5 in Supplementary Material). Mean subclinical elevations were identified in the domains of social problems, thought problems, and attention problems.

Discussion

In this study, we completed deep phenotyping of 17 children and adolescents with MED13L-related disorder and analyzed medical, developmental, and behavioral data of a larger group of up to 67 individuals. These findings support previous case reports of significant cognitive and adaptive delays, but with a larger cohort. Our study is the first to describe detailed speech and language features in MED13L-related disorder. Strikingly, all individuals who completed in-person phenotyping met criteria for a diagnosis of MSD (CAS, childhood dysarthria, or both). Additionally, all individuals who participated in in-person testing exhibited severe receptive and expressive language impairment with commensurate skills in comprehension and production. Almost all individuals (97%) from the full cohort of 67 reported a diagnosis of language disorder during the developmental history interview, with about a third of children described by their caregivers as minimally verbal or nonverbal beyond the age of 4 years. Together, these results suggest that MED13L-related disorder is characterized by pervasive MSD and severe language impairment.

Broad motor developmental delay is another key feature of MED13L-related disorder. Direct testing revealed particularly severe visual motor difficulties in all individuals assessed, with a subset of children unable to participate in direct testing due to their level of impairment. Results of in-person testing are in line with the VABS caregiver report for the larger virtual cohort, in which the group mean standard score in the motor domain fell two S.D.s below the normative mean. Interestingly, caregivers reported high rates of muscle tone abnormalities, coordination difficulties, and movement disorder diagnoses, with two individuals reporting a diagnosis of cerebral palsy. Such findings suggest that MSDs in this group of children are not isolated diagnoses; rather, they occur in the context of globally impaired motor planning and execution.

Other notable findings were the absence of clinically significant internalizing or externalizing behavior problems and relatively low rates of comorbid neuropsychiatric disorders, such as attention deficit/hyperactivity disorder. This is surprising given previous research demonstrating higher rates of behavioral and psychiatric disorders among individuals with intellectual disability compared to the general population [32]. While several subscales of the CBCL did indicate some increased behavioral concerns, there were no identifying behavioral phenotypes. Another unexpected observation was the identification of seven individuals from four unrelated families with P/LP MED13L variants due to presumed germline mosaicism. At least three unrelated individuals with pathogenic MED13L variants because of parental germline mosaicism have been published to date [7–9], bringing the total number to seven families in the literature with presumed or confirmed mosaicism. This is higher than the traditionally reported 1–2% recurrence risk for parents after identification of a de novo variant in their child, but it may be in line with more recent studies suggesting that individualized recurrence risk may be higher for some families [33] Given the seemingly high prevalence of parental mosaicism in MED13L-related disorder, additional research is warranted to systematically assess the prevalence and recurrence risks associated with this condition.

Clinical implications

The high rate of severe MSDs, including a sizable minority of children who are reported by their caregivers as being minimally verbal beyond age four, has clear implications for treatment planning with this population. First, children with known P/LP variants in MED13L would benefit from early referrals to speech therapy to assess their speech, language, and support needs. In our study, 7 children received a genetic diagnosis of MED13L-related disorder before 3 years, the age at which MSDs are typically diagnosed with confidence. There is emerging evidence for preventative speech therapy to proactively mitigate speech and language disorders in children who are known to be at risk for pediatric MSDs due to classic galactosemia [34], supporting the idea that early speech therapy has the potential to improve speech outcomes for young children with a known genetic risk for MSDs. AAC strategies, including the use of speech-generating devices (SGDs), should be carefully considered during the initial assessment. Many children and adolescents in our study, even those who were verbal, used SGDs to improve their functional communication skills. Further, there should be a low threshold for suspicion of MSDs in children with MED13L-related disorder, and evidence-based motor learning techniques should be incorporated into speech therapy when MSDs are suspected. Comprehensive evaluation of fine and gross motor skills is also recommended so that appropriate occupational and physical therapies can be provided.

Limitations and future directions

The standardized vocabulary tests (PPVT-5, EVT-3) that were administered to individuals in person were selected due to their short administration time, simple instructions, wide age range (normed for 2 years, 6 months to 90 + years), and ability to directly compare receptive and expressive skills. However, as single-word vocabulary tests they are limited in scope and do not evaluate other aspects of receptive or expressive language, such as understanding and use of morphology and syntax. More comprehensive language testing for individuals with at least phrase- or sentence-level speech is needed to assess these facets of language. Additionally, differential diagnosis of participants’ speech disorders was made by a single clinician. Although standardized assessments and protocols were utilized to aid with decision-making, these rely on subjective clinical judgment of the presence or absence of specific speech features. Future work focusing on the differential diagnosis of MSDs with this population would benefit from multiple expert raters. Judgments regarding severity, functional impairment, and speech intelligibility would also be useful.

Another limitation of this study is that all participants were ascertained through clinical genetic testing and may not represent the full range of the phenotypic spectrum of MED13L-related disorder. It is possible that more mildly affected individuals with MED13L variants may not have routine access to genetic testing and therefore would not be included in our study. Although beyond the scope of this study, future population studies examining the true prevalence and phenotypic spectrum of MED13L variants would be valuable. Another consideration is that standardized scores for participants who completed in-person testing could be calculated only for the subset of individuals who were able to complete testing. Because individuals with the most severe speech and motor disorders were often unable to participate fully, the average standard scores reported here are likely skewed and may not accurately reflect the lower range of ability in MED13L-related disorder.

Lastly, future studies are needed to elucidate potential effects of MED13L variants on brain development and the relationship between MSDs and neural substrates in this population. One study that examined MED13L gene expression found high expression in both the fetal and adult human brain, particularly in the cerebellum [3]. Because the cerebellum has been implicated in speech motor planning and sequential processing [35, 36], skills required for fluent speech, it is possible that MED13L haploinsufficiency may impair cerebellar structure and/or function, leading to high rates of MSDs in this population. Neuroimaging and gene expression studies, including post-mortem analysis of brain tissue, are needed to test this hypothesis.

Conclusions

MED13L-related disorder is characterized by a high rate of pediatric MSDs, including CAS and childhood dysarthria, that occur in the context of globally impaired motor, language, and cognitive skills. Other behavioral and neuropsychiatric diagnoses are relatively uncommon in this population. These results suggest that children with known P/LP variants in MED13L would benefit from early referrals to speech therapy to assess their speech, language, and support needs.

Abbreviations

MSD

Motor speech disorder

CAS

Childhood apraxia of speech

P/LP

Pathogenic/Likely pathogenic

ACMG

American College of Medical Genetics and Genomics

AMP

Association for Molecular Pathology

GFTA-3

Goldman Fristoe Test of Articulation, Third Edition

DEMSS

Dynamic Evaluation of Motor Speech Skill

ProCAD

Profile of Childhood Apraxia of Speech and Dysarthria

PPVT-5

Peabody Picture Vocabulary Test, Fifth Edition

EVT-3

Expressive Vocabulary Test, Third Edition

DAS-2

Differential Abilities Scale, Second Edition

KBIT-2

Kaufman Brief Intelligence Test, Second Edition

MSEL

Mullen Scales of Early Learning

Beery VMI

Beery-Buktenica Developmental Test of Visual Motor Integration

EHR

Electronic health record

VABS

Vineland Adaptive Behavior Scale

CBCL

Child Behavior Checklist

AAC

Augmentative and alternative communication

Authors’ contributions

Conceptualization and design: MWM, CMT; Data Curation: MWM, ST, LKW, CMT; Funding Acquisition: MWM, CMT; Investigation: MWM, ST, CMT; Visualization: MWM, RIT; Writing-original draft: MWM, ST, SMM, SMT; All authors critically reviewed and revised the manuscript for important intellectual content and approved the final manuscript.

Funding

This study was supported by grants from the Simons Foundation Autism Research Initiative-Simons Searchlight (1003050), the National Institute of Mental Health (U01MH119705), and the National Institute of Child Health and Human Development (R01HD104938).

Data availability

Deidentified participant data from Simons Searchlight use din this study can be obtained by request (http://base.sfari.org). Reuse of data is permitted on submission of a data request. Additional data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Written informed consent for study participation was obtained from all participants or their legal representatives. Ethics approval for this study was obtained from the Geisinger Institutional Review Board (#00008345) under protocol #2013–0446. This research was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.