TRPV4 neuromuscular disease registry highlights bulbar, skeletal and proximal limb manifestations

Abstract

Dominant missense mutations of the calcium-permeable cation channel TRPV4 cause Charcot-Marie-Tooth disease (CMT) type 2C and two forms of distal spinal muscular atrophy. These conditions are collectively referred to as TRPV4-related neuromuscular disease and share features of motor greater than sensory dysfunction and frequent vocal fold weakness. Pathogenic variants lead to gain of ion channel function that can be rescued by TRPV4 antagonists in cellular and animal models. As small molecule TRPV4 antagonists have proven safe in trials for other disease indications, channel inhibition is a promising therapeutic strategy for TRPV4 patients. However, the current knowledge of the clinical features and natural history of TRPV4-related neuromuscular disease is insufficient to enable rational clinical trial design. To address these issues, we developed a TRPV4 patient database and administered a TRPV4-specific patient questionnaire.

Here, we report demographic and clinical information, including CMT Examination Scores (CMTES), from 68 patients with known pathogenic TRPV4 variants, 40 of whom also completed the TRPV4 patient questionnaire. TRPV4 patients showed a bimodal age of onset, with the largest peak occurring in the first 2 years of life. Compared to CMT type 1A (CMT1A) patients, TRPV4 patients showed distinct symptoms and signs, manifesting more ambulatory difficulties and more frequent involvement of proximal arm and leg muscles. Although patients reported fewer sensory symptoms, sensory dysfunction was often detected clinically. Many patients were affected by vocal fold weakness (55%) and shortness of breath (55%), and 11% required ventilatory support. Skeletal abnormalities were common, including scoliosis (64%), arthrogryposis (33%) and foot deformities. Strikingly, patients with infantile onset of disease showed less sensory involvement and less progression of symptoms.

These results highlight distinctive clinical features in TRPV4 patients, including motor-predominant disease, proximal arm and leg weakness, severe ambulatory difficulties, vocal fold weakness, respiratory dysfunction and skeletal involvement. In addition, patients with infantile onset of disease appeared to have a distinct phenotype with less apparent disease progression based on CMTES. These collective observations indicate that clinical trial design for TRPV4-related neuromuscular disease should include outcome measures that reliably capture non-length dependent motor dysfunction, vocal fold weakness and respiratory disease.

TRPV4-related neuropathy is potentially treatable with antagonist drugs, but clinical trial design requires a better understanding of disease symptoms and how they change over time. Kosmanopoulos et al. studied a large group of TRPV4 patients to identify core disease features and potential clinical trial outcome measures.

Introduction

Charcot-Marie-Tooth disease (CMT), also known as hereditary motor and sensory neuropathy, comprises a group of hereditary peripheral neuropathies that together represent the most common inherited neurological condition worldwide with a prevalence of approximately 1 in 2500.^1,2 CMT is characterized by progressive motor and sensory dysfunction, most often beginning distally in the legs and later progressing to involve the hands and arms.³ Although the genetic landscape of CMT has dramatically expanded over the past three decades, with identification of now more than 100 causative genes, there are currently no approved disease-modifying treatments for any form of CMT.^4,5 Autosomal dominant missense mutations of TRPV4 (transient receptor potential vanilloid 4) have been identified as the cause of CMT type 2C (CMT2C)⁶ and two recognized forms of spinal muscular atrophy (SMA), scapuloperoneal SMA (SPSMA)^7,8 and congenital distal SMA (CDSMA).⁸ Whereas CMT most often leads to length-dependent muscle weakness and sensory abnormalities, SMA is a motor neuronopathy that affects proximal and distal arm and leg muscles.⁹ All three disease entities caused by TRPV4 mutations result in motor dysfunction and are distinguished clinically by the presence or absence of sensory involvement (present in CMT2C and absent in SPSMA/CDSMA), the pattern of arm weakness (distal in CMT2C/CDSMA and proximal in SPSMA) and the age of onset (congenital in CDSMA).¹⁰ However, as the validity of these clinical labels has not been evaluated in a large cohort, there may exist overlap among these three disease presentations. In addition, other dominant TRPV4 mutations cause distinct diseases of connective tissue and bone, including a spectrum of skeletal dysplasias and a form of hereditary osteoarthropathy.^10-13 The non-neurological manifestations of TRPV4-related neuromuscular disease have not been systematically studied.

TRPV4 encodes a plasma membrane-expressed non-selective ion channel that is preferentially permeable to calcium.¹⁴ Since identification of TRPV4 mutations as a cause of hereditary neuromuscular disease, numerous in vitro and in vivo studies have demonstrated that disease-associated mutations lead to gain of ion channel function with increased calcium influx and subsequent cytotoxicity.^15-19 Furthermore, toxicity in cellular, fly and mouse models of disease can be rescued with selective small molecule TRPV4 antagonists.^6,7,15,20 As increased activity of TRPV4 has been linked to disease pathogenesis, and no loss-of-function mutations in TRPV4 have been shown to cause human disease, inhibition of TRPV4 ion channel activity is a promising therapeutic strategy for patients with TRPV4-related neuromuscular disease.²¹ However, incomplete knowledge of the clinical features and disease progression as well as ambiguity as to whether TRPV4-related neuromuscular disease represents a peripheral neuropathy (as in CMT) and/or motor neuronopathy (as in SMA) are barriers to clinical trial design. The distinction between CMT and SMA is important as clinical outcome measures and efficacious treatments for SMA due to SMN1 deficiency have been developed and proven in large clinical trials,²² but responsive outcome measures have yet to be validated in a clinical treatment trial for CMT.

Available data for TRPV4 patients have come from studies of small cohorts, families and individual patients.^6-8,23-25 These studies have revealed a wide range of disease manifestations and marked heterogeneity in symptoms, severity and age of onset, even among family members with the same TRPV4 mutation.¹⁰ In fact, the same TRPV4 mutation can cause CMT or SMA presentations, even within the same family. The prevalence of TRPV4-related neuromuscular disease has not been clearly established, with estimates of ∼1%–3.5% of CMT2/hereditary motor neuropathy.^26-28 Typical features include upper and lower extremity weakness, vocal fold weakness and skeletal deformities.¹⁰ Additional recognized clinical features include scapular winging, diaphragmatic weakness, sensorineural hearing loss, absence of specific brainstem-innervated muscles, arthrogryposis, sleep apnoea and tethered spinal cord.^6-8,23-25,29 Despite the variability among patients, an important and defining feature of TRPV4 neuropathy is vocal fold weakness, which can be severe and life threatening.¹⁰ However, the continued identification of new clinical features suggests that there may be other unrecognized aspects of the TRPV4-related neuromuscular disease phenotypic spectrum.

To define the clinical features of TRPV4-related neuromuscular disease and aid in identification of responsive clinical outcome measures, we sought to analyse a large cohort of patients with respect to core clinical features, longitudinal disease progression and patient-reported outcomes. To this end, we enrolled patients from across the world into a TRPV4 patient registry as part of the Inherited Neuropathy Consortium (INC). To supplement these data and specifically interrogate unusual disease features that are not traditionally captured by CMT evaluations, we also developed a TRPV4-specific patient questionnaire. Notable findings from our analysis include a high burden of proximal lower extremity weakness, frequent bulbar involvement, often asymptomatic sensory dysfunction, frequent skeletal manifestations and differences in disease progression based on age of disease onset. These findings will help guide multidisciplinary care for TRPV4 patients as well as inform further evaluation of clinical outcome measures and surrogate biomarkers for future clinical trials.

Materials and methods

Study design and patient enrolment

Patients included in this study were enrolled in the INCRDCRN 6601 natural history protocol (registered at ClinicalTrials.gov NCT01193075), which gained ethical approval from the institutional review boards and research ethics committees of the participating centres in the USA, UK and Italy. All patients or their guardians signed the relevant consent/assent forms. Only patients with personal genetic testing results or results from a first degree relative and symptoms consistent with TRPV4-related neuromuscular disease were considered for analysis. A total of 66 patients were evaluated at one of the 12 INC centres between 2009 and 2023, and 29 patients were evaluated through a video visit. Patients were from 74 families (13 families with two members enrolled, four families with three members enrolled and 57 patients with no other family members enrolled). Antecedent clinical data were collected retrospectively from the patient history. Longitudinal clinical follow-up data was collected prospectively during annual visits.

Clinical evaluations

Patients were evaluated using the Rasch-weighted Charcot-Marie-Tooth Neuropathy Score (CMTNS)³⁰ or its subcomponents, the Charcot-Marie-Tooth Examination Score (CMTES) or the Charcot-Marie-Tooth Symptom Score (CMTSS), which have been validated in natural history studies of other forms of CMT.^31-34 A subset of children under 10 years of age were evaluated using the CMT Pediatric Scale (CMTPedS).³⁵ Patients evaluated by video visit underwent assessment using the Virtual Charcot-Marie-Tooth Examination Score (vCMTES).³⁶ The CMTNS is a validated 9-item composite score of CMT disease severity that includes sensory and motor symptoms, sensory and motor neurological examination and electrophysiology. It is scored on a range of 0–36 with higher scores indicating more severe disability. The CMTES is a seven-item subscore of the CMTNS that excludes electrophysiology measures, which are not routinely repeated on follow-up clinic visits. The CMTSS is a three-item subscore of the CMTNS that excludes electrophysiology and neurological examination components. The vCMTES is a seven-item modified version of the CMTES that replaces pinprick and vibration items with light touch and position sense, as these evaluations can be performed by an assistant and observed remotely by video link. CMTPedS is a validated 11-item scale that incorporates upper and lower extremity functional measures of strength, dexterity, sensation, gait, balance, power and endurance. It is scored from 0–44 with higher scores indicating more severe disability. All outcome measures were obtained by trained evaluators.

TRPV4 patient questionnaire

We created a TRPV4 patient questionnaire to assess for the presence and severity of various signs and symptoms associated with CMT as well as those specifically associated with TRPV4-related neuromuscular disease. The questionnaire was designed to complement the clinical dataset captured by enrolment in the INCRDCRN 6601 natural history protocol. Briefly, questions addressed age of onset, age of diagnosis, bulbar and respiratory symptoms, weakness, sensory loss, pain, upper and lower limb function, skeletal disease, skin problems, urinary problems, autonomic dysfunction and the most bothersome and problematic symptoms. For several items on the questionnaire, the patients were asked to report the age of onset of signs or symptoms, severity or frequency of symptoms and/or how they were changing over time. We also included questions relating to possible underrecognized or unrecognized features, some of which have only recently been reported. The adult version (for adults and children 7 years of age and older) consisted of 33 items, and the paediatric version (for children under 7 years of age) consisted of 24 items. The paediatric questionnaire omitted questions relating to symptom changes over time. The questionnaire was administered for each patient only once at initial enrolment and was not readministered at follow-up visits. The full questionnaires are included as Supplementary material.

Comparison between TRPV4-related neuromuscular disease and CMT1A

Data from CMT1A patients available within the INCRDCRN 6601 natural history study were utilized to compare the phenotypic features of TRPV4-related neuromuscular disease with CMT1A. Baseline clinical data from CMT1A patients (n = 1680) enrolled between and 2009 and 2021 was included for analysis. Analysis was limited to age at enrolment and specific components of the CMTES.

Statistical analysis

Patient demographics, clinical features, CMTNS, CMTES, CMTSS and questionnaire responses were analysed using descriptive methods in Prism version 10 (GraphPad) and Stata version 17 (StataCorp). Simple linear regression was used to analyse CMTES, CMTNS, radial sensory responses and ulnar motor responses versus age. For analysis of infantile onset and non-infantile onset questionnaire results, ordinal variables were recoded as numbers in terms of severity and compared by Wilcoxon rank, and categorical variables were compared using chi-squared testing. Two-tailed P-values of <0.05 were considered statistically significant. Rasch analysis was performed using Winstep software (version 5.7.0.1) to further explore psychometric properties and construct validity of question 13 on the TRPV4 questionnaire as a potentially sensitive disease severity survey. This analysis included item-person targeting (how well items cover the range of disease severity), item fitting (how well items add up to form a linear measure), patient fitting (how well individual patients follow the same pattern of involvement) and category response functioning and weighting (how well scale items can differentiate between different levels of disease severity). Principal component analysis of fit residuals was performed to test unidimensionality of the outcome.

Results

TRPV4-related neuromuscular disease is motor predominant with bimodal age of onset

A total of 95 patients with TRPV4 variants were enrolled by 12 INC centres across the USA, UK and Italy. Within the cohort, 68 patients had TRPV4 variants that were classified as pathogenic or likely pathogenic or had in vitro evidence of pathogenicity.^19,37 The remaining 27 patients had variants that were classified as benign, likely benign, or of uncertain significance and were excluded from further analysis. The demographics and clinical characteristics of the remaining 68 patients at their first visit are summarized in Tables 1 and 2. Ambulatory difficulties were common, with 79% of patients having difficulty walking, 79% having difficulty with balance and 32% requiring at least intermittent use of a wheelchair. Foot deformities were present in 73%, including pes cavus in 33% and hammertoes in 22% of those with foot deformities. Foot surgery had been performed in 48% of patients, including tendon transfer, Achilles tendon lengthening, or ankle joint fusion. Difficulties with hand function were also common, with 49% of patients reporting difficulties with buttons, zippers, fasteners or bottles and 23% reporting difficulties with using eating utensils. Hearing loss was reported in 27%. Sensory symptoms were less prominent than motor symptoms, with 17% reporting burning or tingling in the hands or feet and 38% reporting decreased ability to feel.

Table 1.

TRPV4 patient demographics

Demographics	n (%)	Total
Total number of patients	68	68
Median age at baseline visit (IQR)	39 (18–60)	60
Male	29 (44)	66
Female	37 (56)	66
Race	–	–
White	56 (82)	68
Asian	6 (8.8)	68
Black/African American	1 (1.5)	68
Mixed (identifies as Black and White)	1 (1.5)	68
Other	1 (1.5)	68
Not reported	3 (4.4)	68

IQR = interquartile range.

Table 2.

TRPV4 patient baseline clinical features

Baseline clinical characteristics	n (%)	Total
Walked before 15 months	35 (63)	56
Difficulty walking	54 (79)	68
Difficulty walking median age of onset (IQR)	2 (2–28)	43
Hip dysplasia	12 (18)	66
Difficulties with balance	53 (79)	67
Difficulties with balance median age of onset (IQR)	4 (4–42)	45
Orthotic aids	33 (49)	67
Shoe inserts	6 (18)	33
Ankle-foot orthotics	22 (67)	33
Other	6 (18)	33
Walking support needs	22 (34	64
Wheelchair use required	21 (32)	66
Foot deformity	49 (73)	67
Pes planus	9 (18)	49
Pes cavus	16 (33)	49
Hammertoes	11 (22)	49
Other	21 (43)	49
Foot surgery	32 (49)	66
Tendon transfer	7 (22)	32
Achilles tendon lengthening	7 (22)	32
Ankle joint fusion	7 (22)	32
Osteotomy	6 (19)	32
Toes straightened	3 (9.4)	32
Other	21 (66)	32
Difficulties with eating utensils	15 (23)	65
Difficulties with buttons/zippers/fasteners/bottles	33 (49)	67
Hand surgery	7 (11)	65
Hand surgery involving bone	0 (0)	7
Hand surgery involving soft tissue	4 (57)	7
Tendon transfer procedure	0 (0)	7
Carpal tunnel procedure	5 (71)	7
Other procedures	4 (57)	7
Partial hearing loss	18 (27)	67
Burning or tingling in feet or hands or other paresthesia	11 (17)	63
Decreased ability to feel	24 (38)	63
Ulcerations in feet or hands	3 (4.6)	65
Amputations feet or hands	0 (0)	66
Arthritic-like (aching) pain	30 (49)	61
Ache in ankles	14 (47)	30
Ache in knees	16 (53)	30
Ache in hips	13 (43)	30
Ache in other	19 (63)	30
Scoliosis	26 (39)	66
Scoliosis surgery	10 (38)	26
Scoliosis bracing	3 (12)	26
Sural nerve biopsy	3 (5.0)	60

IQR = interquartile range.

We next analysed CMT severity scores at baseline enrolment for the 68 patients with pathogenic TRPV4 variants. Of these, 61 patients had recorded baseline CMT scores; nine were <10 years old, four were 11–18 years old, and the remaining 48 were adults >18 years of age. Mean Rasch-weighted CMT scores for the entire cohort were 4.7 for CMTSS (range 0–10, n = 61), 11 for CMTES (range 0–22, n = 61) and 13 for CMTNS (range 1–22, n = 34) (variable n for each analysis reflects missing data for some evaluations). The mean CMTES score was in the moderate range of clinical severity (mild 0–9, moderate 10–18 and severe 19–32).³¹ Males and females in the cohort had similar average ages (males 39 years versus females 48 years) and CMTES scores (males 13 versus females 12). There was a low correlation (R² = 0.066) between age and CMTES at baseline evaluation (Fig. 1A), indicating minimal progressive disease burden over time. By comparison, this correlation was stronger in patients with CMT2A (R² = 0.196) and males with CMTX1 (R² = 0.335).^32,33 We found a similar low correlation between CMTSS and age (R² = 0.062; Supplementary Fig. 1A) and CMTNS and age (R² = 0.129; Supplementary Fig. 1B). Given reduced reliability of CMTNS in children, we also analysed CMTSS, CMTES and CMTNS after excluding patients less than 10 years of age. Results were similar, but correlations of age versus CMT scores were somewhat more robust (Supplementary Fig. 1C–E). Together, these results suggest that TRPV4-related neuromuscular disease progresses slowly or, alternatively, that CMTSS, CMTES and CMTNS do not adequately capture disease features that change over time. We also analysed whether sensory or motor nerve conduction studies showed correlation with age. As expected, nerve conduction studies showed variable reductions of radial SNAP and ulnar CMAP amplitudes with preserved conduction velocities (mean ulnar motor conduction velocity 57.2 ± 7.2, n = 14), consistent with loss of large, myelinated axons. Radial SNAP and ulnar CMAP showed non-significant correlations with age (radial SNAP: slope = −0.027, R² = 0.0025, P = 0.88, n = 12; ulnar CMAP: slope = −0.044, R² = 0.11, P = 0.26, n = 14) (Supplementary Fig. 1F and G), suggesting that these measures may not capture changes in disease severity over time.

Figure 1.

TRPV4-related neuromuscular disease is motor-predominant. (A) Patient age at baseline registration in the patient registry plotted versus Charcot-Marie-Tooth Examination Score (CMTES). (B) Histogram of patient questionnaire responses for age of symptom onset. (C) Histogram of patient questionnaire responses for age of diagnosis. (D and E) Graphs of percentages of patients reporting weakness of regions of the arms (D) and legs (E) from the TRPV4 patient questionnaire. (F) Graph of percentages of patients reporting difficulties with various arm and leg functions. (G and H) Graphs of percentages of patients reporting sensory symptoms (G) or pain (H) in specific body regions.

Given that patients with TRPV4 mutations frequently have clinical features that are unusual in other forms of CMT, we also created a TRPV4-specific patient questionnaire to complement the INC patient registry data. The questionnaire was completed for 40 patients with pathogenic mutations (34 adults and six children). Patient-reported age of symptom onset showed a bimodal distribution, with 54% having symptom onset within the first 2 years of life (Fig. 1B). There was a second, smaller peak at 34–48 years of age. Age at diagnosis was markedly shifted compared to age of onset, with the largest peak at 34–48 years of age and smaller peaks at 0–2 and 49–64 years of age (Fig. 1C). Thus, many patients experienced a significant delay between symptom onset and genetic discovery of a pathogenic TRPV4 mutation (TRPV4 mutations were first identified in 2010 as a cause of neuromuscular disease^6-8). Patients frequently reported widely distributed weakness of the arms and legs that affected proximal and distal muscles (Fig. 1D and E). Most patients reported little or no difficulty with most upper extremity activities, such as writing, opening a jar, or dressing, whereas lower extremity activities were frequently affected, with the majority of patients reporting inability to walk on toes (73%), walk 1 mile (56%), hike (62%) or run (78%) (Fig. 1F). In contrast, sensory symptoms and pain were less prominent, with most patients reporting pain occurring less than half of all days within specific body regions (Fig. 1G and H). Collectively, these data highlight that TRPV4-related neuromuscular disease typically is motor-predominant.

We used Rasch analysis to explore psychometric properties of the questionnaire, focusing on question 13 as a potential functional outcome measure. We analysed item-person targeting, stability (fitting) and category response functioning. Overall, there was no significant floor or ceiling effect (Supplementary Fig. 2A). Leg function-related items appeared to be sensitive to mildly affected patients, whereas arm function-related items appeared more relevant to severely affected patients. All leg function-related items fit to the expectation of the Rasch model, whereas three arm function-related items (opening jars, lifting arms and getting dressed) were slightly misfitting (Supplementary Fig. 2A). The impact of misfitting items on the overall unidimensionality of the outcome was negligible. Category response functioning analysis demonstrated that all items had ordered category response functioning and scoring, with patients with more disability responding higher in all categories (Supplementary Fig. 2B). There was a strong correlation between total calculated functional score on question 13 and CMTES (r = 0.8, P < 0.001), showing a strong convergent validity for this question as an outcome measure.

TRPV4 clinical features are distinct from CMT1A

The patient demographics and questionnaire responses suggested a higher frequency of ambulatory difficulties and proximal arm and leg weakness in patients with TRPV4 mutations as compared to other forms of CMT. Thus, we compared the motor and sensory features in the 68 TRPV4 patients to those with CMT1A, who account for nearly 50% of all CMT cases and have the prototypical CMT manifestations. The CMT1A cohort used in this study was from the INCRDCRN 6601 registry and has been described previously.³¹ Despite similar age distributions in the two cohorts (Fig. 2A), baseline evaluations of TRPV4 patients demonstrated more severe leg weakness and ambulatory difficulties, and a higher use of walking aids (10% CMT1A versus 21% TRPV4) and a wheelchair (1% CMT1A versus 19% TRPV4) (Fig. 2B). These patient-reported symptoms were consistent with clinical exams showing that 40% of TRPV4 patients had proximal leg weakness versus 4% of CMT1A patients (Fig. 2C). Proximal arm weakness was more common in TRPV4 patients based on both patient report and clinical examination (Fig. 2D and E). In contrast, sensory symptoms were less frequent in TRPV4 patients, with 68% reporting no sensory symptoms (Fig. 2F). However, clinical examination showed diminished pinprick sensation in 47% of TRPV4 patients and diminished vibration sensation at similar rates to CMT1A patients (84% TRPV4% versus 83% CMT1A) (Fig. 2G and H). Thus, subclinical sensory involvement is common in TRPV4 patients. Overall, these results suggest that motor dysfunction in TRPV4 neuropathy often follows a non-length-dependent pattern that is distinct from typical CMT1A, and that asymptomatic sensory involvement may be more common than previously appreciated.

Figure 2.

TRPV4 clinical features are distinct from Charcot-Marie-Tooth disease type 1A (CMT1A). (A) Box and whisker plot of average age at baseline registration for CMT1A and TRPV4 patients. Boxes represent the 25%–75% percentile, and bars represent maximum and minimum values. (B and C) Histogram of leg symptoms (B) and leg strength examination (C) in CMT1A versus TRPV4 patients. (D and E) Histogram of arm symptoms (D) and arm strength examination (E) in CMT1A versus TRPV4 patients. (F) Histogram of sensory symptoms in CMT1A versus TRPV4 patients. (G and H) Histogram of vibration sensation (G) and pinprick sensation (H) in CMT1A versus TRPV4 patients.

Charcot-Marie-Tooth disease exam scores demonstrate minimal disease progression

As the longitudinal course of TRPV4-related neuromuscular disease remains largely unknown, we investigated whether disease progression could be detected in 13 patients who had undergone at least two repeated neurological examinations and CMTES assessments (follow-up duration range 2–9 years). Analysing the change in CMTES from baseline in this cohort showed an average change of +0.81 in CMTES per year (R² = 0.31, P < 0.0001) (Fig. 3A and Supplementary Table 1). The progression for individual patients was highly variable, with some showing little to no change in CMTES over time (Fig. 3B and C), and others showing progressive worsening (Fig. 3D–F). The variability in CMTES over repeated visits in several patients (Fig. 3D and E) resulted in marked deviations in the average yearly change in CMTES. The variable longitudinal change likely reflects the low number of patients studied and may also indicate challenges in applying the CMTES to TRPV4 patients who often have non-length-dependent clinical signs and symptoms. CMTPedS scores were available for three patients, and longitudinal scores were available for two patients. Severity scores were in the mild to moderate range, with one patient showing worsening over time and one patient showing stable disability (Fig. 3G).

Figure 3.

TRPV4-related neuromuscular disease progression is variable. (A) Average change in baseline Charcot-Marie-Tooth Examination Score (CMTES) versus year of follow-up for 13 patients from the patient registry. Error bars represent the standard error of the mean. (B–F) Longitudinal CMTES scores over repeated evaluations for individual patients from the patient registry. (G) Age versus CMT Pediatric Scale (CMTPedS) scores for three paediatric patients.

Patient quality of life is driven by motor and bulbar symptoms

We analysed questionnaire data to assess additional features of TRPV4-related neuromuscular disease. In keeping with the known of vocal fold dysfunction in TRPV4 patients, we found that 55% of all patients reported being diagnosed with vocal fold weakness, and the percentage was even higher (85%) of those who reported having undergone otorhinolaryngology evaluation (Fig. 4A). About half of TRPV4 patients reported shortness of breath (55%) and hoarseness (45%) (Fig. 4A and B). There was also a high frequency of inspiratory stridor (40%), need for respiratory support with CPAP (10%) or BiPAP (23%), and over 10% of patients reported requiring ventilatory support and/or having undergone tracheostomy (Fig. 4A and B).

Figure 4.

Patient experience is driven by motor and bulbar symptoms. (A) Graph of percentages of patient responses to questions regarding vocal fold and bulbar function and assessments. (B) Graph of percentages of patient responses to questions regarding respiratory function and assessments. (C) Graph of percentages of patient responses to questions regarding skeletal disease. (D) Graph of percentages of patient responses to questions regarding urinary symptoms. (E and F) Pie charts showing the percentages of patient responses to the question of which symptom is the most bothersome (E) and which symptom has the greatest impact on daily life (F).

In addition to neuromuscular disease, distinct dominant mutations in TRPV4 also cause a spectrum of skeletal disorders, including a heterogeneous group of skeletal dysplasias with features of short stature, vertebral body flattening (platyspondyly), scoliosis, and shortened digits (brachydactyly)^14,38,39 and familial digital arthropathy brachydactyly (FDAB), a progressive, painful and deforming osteoarthropathy affecting the hands and feet.¹³ While the neuromuscular and skeletal phenotypes were once considered to be distinct and non-overlapping, patients with combined neuromuscular and skeletal dysplasia features are increasingly recognized.¹⁹ Consistent with overlap of neuromuscular and skeletal disease with TRPV4 mutations, questionnaire responses revealed many common skeletal manifestations (Fig. 4C), including scoliosis (64%), clubfoot (45%), arthrogryposis (33%), ‘skeletal dysplasia’ (20%) and tethered spinal cord (7.5%). This latter finding is striking given that tethered spinal cord was only recently reported to occur in this patient population.^19,29

We also assessed for the possibility of other unrecognized symptoms in TRPV4 patients. Given the known high expression of TRPV4 in the bladder and urinary dysfunction noted in TRPV4 knockout mice,⁴⁰ we also asked patients about urinary symptoms. Difficulty with urination was reported by 45% of patients, with incontinence occurring in 26% and urinary frequency occurring in 29% of patients (Fig. 4D). We also screened patients for skin symptoms based on known high expression of TRPV4 in skin and a recent report of skin problems in a family with a W785C TRPV4 mutation.^41,42 Nearly 50% of patients denied any skin problems, while 35% reported scaliness or dryness (Supplementary Fig. 3A). Autonomic symptoms of tachycardia (26%) and orthostatic dizziness (41%) were also present in some patients (Supplementary Fig. 3B).

To assess which of the myriad disease manifestations were the most important to patient quality of life, we asked patients to report the symptom that ‘bothers you the most (is the most unpleasant symptom)’ and the symptom that ‘has the greatest impact on daily life (interferes with daily activities the most)'. With respect to the most bothersome symptom, speech and breathing problems accounted for 22% of responses, lower extremity weakness and gait difficulties 21% and weakness of the upper extremities 9% (Fig. 4E). For the symptom with the greatest impact on daily life, walking difficulties were cited by 32% and hand weakness by 17% (Fig. 4F). In aggregate, weakness of either the upper or lower extremities accounted for 74% of responses for the symptom with the greatest impact on daily life. These data indicate that symptoms related to vocal fold dysfunction and extremity weakness, particularly of the legs, are primary determinants of reduced quality of life for many TRPV4 patients.

TRPV4 genotype-phenotype correlation and subtype analysis

We next analysed the genetic variants within the overall patient registry cohort and in those that completed the questionnaire. Most mutations that cause TRPV4-related neuromuscular disease occur within the cytosolic ankyrin repeat domain (ARD, residues 149–386) and affect conserved arginine residues that mediate protein-protein interactions.^43-45 Variants within the ARD accounted for 94% of all neuropathogenic mutations, and the R269H mutation was the most common mutant overall (Fig. 5A). The mutation frequency was similar for those that completed the questionnaire (Supplementary Fig. 4A). There were no clear genotype-phenotype correlations between specific pathogenic variants and CMTES at baseline examination (Fig. 5B), although the age of symptom onset was lowest in patients with the R269H mutation (Fig. 5C). We then analysed each pathogenic variant with respect to CMTES at time of baseline evaluation. The correlation of age and severity was weakest for the R269H and R315W mutations (Supplementary Fig. 4B and C), and correlations were more robust with more positive slopes for the remaining pathogenic variants within the ARD (Supplementary Fig. 4D–G).

Figure 5.

TRPV4 genotype-phenotype evaluation. (A) Pie chart showing the frequency of specific TRPV4 mutations within patients in the registry. (B) Graph showing average Charcot-Marie-Tooth Examination Score (CMTES) at baseline registration for specific TRPV4 mutations. (C) Graph showing average age of onset for specific TRPV4 mutations. (D–G) Graphs of CMTES arm exam score versus leg exam score (D), CMTES arm exam score versus patient reported sensory symptoms (E), CMTES arm exam score versus vibration score (F) and CMTES arm exam score versus pinprick score (G). Red bar denotes patients with proximal arm weakness. One-way ANOVA, ***P < 0.001, comparing patients weak above the elbows to other groups. (H–K) Graphs comparing patients with or without significant shoulder weakness with respect to frequency of arm and leg weakness (H), difficulties with activities (I), sensory loss (J) and pain (K). Error bars represent the standard error of the mean.

We next sought to determine whether we could discriminate different patterns of motor and sensory involvement that could differentiate between CMT2C and SPSMA phenotypes. Given that SPSMA is largely defined by proximal shoulder girdle weakness and lack of sensory involvement, we sorted patients by the degree of proximal arm and proximal leg weakness on baseline neurological examination and then analysed other clinical features in these groups. Notably, those with more severe proximal arm weakness also tended to have more severe and proximal leg weakness (Fig. 5D) and vice versa (Supplementary Fig. 4H). In contrast to what would be expected if proximal arm weakness were an indicator of the SPSMA phenotype, we found no correlation of proximal arm weakness with the presence of sensory symptoms or objective abnormalities of vibration and pinprick sensation (Fig. 5E–G). We also saw no correlation of proximal leg weakness with sensory dysfunction (Supplementary Fig. 4H–K). To complement this analysis, we dichotomized questionnaire respondents based on significant difficulties lifting the arms over the head (either ‘significant difficulty’ or ‘unable to do’), which we reasoned should be an indicator of proximal arm weakness and shoulder girdle dysfunction that are characteristic of SPSMA. Consistent with the above analysis of patient registry data, we found that patients who reported significant difficulty lifting the arms had more severe weakness of additional muscle groups in the arms and legs and more severe difficulties with activities involving the limbs (Fig. 5H and I). We also found that patients with significant shoulder weakness reported more sensory loss in the fingers/hands and toes/feet than those without shoulder weakness (Fig. 5J), which is the opposite of what would be expected for patients with SPSMA in whom sensory involvement is thought to be minimal. Pain was not different between the two groups (Fig. 5K). Together, these analyses suggest that the presence of proximal arm and leg weakness does not reliably predict a motor-exclusive phenotype and argues against shoulder girdle weakness as a marker of the SPSMA subtype of TRPV4-related neuromuscular disease.

Patients with infantile onset show less sensory involvement and minimal disease progression

Given prior observations suggesting that patients with congenital onset disease, often referred to as CDSMA, appear to have a largely static clinical course^8,10,46,47 we sought to determine whether patient report of symptom progression in the questionnaire differed between infantile onset (here defined as ≤2 years of age) and non-infantile onset (>2 years of age) patients. The average age of the two groups at baseline evaluation were 30.5 years for the infantile onset patients and 48.0 years for the non-infantile onset patients. Notably, non-infantile onset patients reported a higher frequency of hoarseness, stridor and shortness of breath, and higher frequencies of progression of symptoms (Fig. 6A). While the frequency of upper and lower extremity weakness was not markedly different in the two groups, a higher fraction of non-infantile onset patients reported worsening of arm, hand, lower leg and foot strength (Fig. 6B and C). They also reported a higher rate of worsening with respect to arm and leg function, with statistically significant increases in worsening of walking on toes, walking a mile and hiking (Fig. 6D). Strikingly, sensory changes and pain were significantly more common in non-infantile onset patients as compared to infantile onset (Fig. 6E and F). Overall, these differences suggest that infantile onset patients may have a distinct clinical phenotype and a slower rate of progression. In light of these findings, we reanalysed the correlation of age versus baseline CMTES in the two cohorts. This analysis showed clear correlation between age and baseline CMTES in non-infantile onset patients (R² = 0.38, slope 0.29, P = 0.008) and virtually no correlation in infantile onset patients (R² = 0.018, slope 0.033, P = 0.57) (Fig. 6G). These findings underscore heterogeneity in disease severity and progression and suggest that congenital/infantile-onset patients may have a distinct clinical phenotype.

Figure 6.

Patients with infantile-onset disease have less sensory involvement and slower progression. (A–C) Graphs of percentages of infantile-onset and non-infantile onset patient responses to questions regarding vocal fold symptoms and progression (A), arm weakness and progression (B) and leg weakness and progression (C). The fraction of each patient cohort with a given symptom is represented by the full height of the coloured bars, and the fraction with progression by the darker portion of the bar. (D) Graph of percentages of infantile-onset and non-infantile onset patients reporting difficulties with various arm and leg functions. Chi-squared test, *P < 0.05 comparing progressing fractions of each group (dark blue and dark red bars). (E and F) Graphs of percentages of infantile-onset and non-infantile onset patients reporting sensory changes (E) and pain (F) in specific body regions. Chi-squared test, *P < 0.05 comparing total fractions of each group (full height of blue and red bars). (G) Charcot-Marie-Tooth Examination Score (CMTES) versus age at registration for infantile-onset and non-infantile onset patients.

Discussion

TRPV4-related neuromuscular disease has emerged as a potentially treatable condition.²⁰ However, detailed cross-sectional phenotypic data and longitudinal clinical assessments to guide clinical trial design have been lacking. Furthermore, the heterogeneity of disease severity creates challenges for choosing clinical outcome measures. To begin to address these barriers, we developed and analysed a TRPV4 patient registry and supplemented the available data with a TRPV4 patient questionnaire. Our patient cohort represents the largest cross-sectional and longitudinal analysis of TRPV4-related neuromuscular disease patients to date. Analysis of these data demonstrate a series of core clinical features of TRPV4-related neuromuscular disease, including motor predominant signs and symptoms, non-length-dependent patterns of weakness in the extremities, frequent vocal fold and bulbar dysfunction and a high rate of skeletal involvement. The finding of non-length dependent limb weakness suggests a component of neuronopathy with pathology localized to the motor neuron cell body. We also found a striking bimodal age of disease onset with the largest peak occurring within the first 2 years of life. Importantly, our data demonstrate that weakness, particularly involving the legs, and speech and respiratory symptoms are the strongest contributors to patient-reported quality of life and should be considered in clinical trial design.

Diverse disease manifestations require multidisciplinary care

Our data highlight the diverse, multisystem manifestations of TRPV4-related neuromuscular disease. In addition to sensory and motor dysfunction, patients frequently reported speech and respiratory difficulties, which could be attributable to a combination of vocal cord dysfunction, diaphragmatic weakness and potentially other abnormalities of the airway. Respiratory dysfunction was severe in a subset of patients and required tracheostomy and/or ventilatory support. While not typically considered a prominent aspect of TRPV4-related neuromuscular disease, over 30% of patients reported difficulties swallowing. In addition, use of CPAP in many patients suggests the possibility of sleep-disordered breathing. Skeletal disease was common, manifesting most commonly as scoliosis, foot deformities and joint contractures. These findings suggest that optimal care for TRPV4 patients requires a tailored, multidisciplinary approach with routine involvement of neuromuscular neurology, otolaryngology, speech pathology, pulmonology and orthopaedics. Referral to additional medical specialists should be considered as needed, such as neurosurgery for evaluation and management of tethered spinal cord and sleep medicine for sleep-disordered breathing. Together, our results also highlight the need for clinicians to be comprehensive and vigilant in screening for atypical disease manifestations and involving a multidisciplinary clinical team as appropriate.

Clinical heterogeneity and subtype analysis

Prior clinical evaluations of TRPV4 patients and families have led to division of patients into three diagnostic entities: CMT2C, SPSMA and CDSMA. These distinctions have largely been made based on the presence of sensory symptoms in CMT2C, the absence of sensory symptoms in SPSMA and CDSMA and congenital onset in CDSMA. However, the existence of these disease subtypes and their clinical definitions have not been validated in large patient cohorts. To determine whether these disease subtypes could be discriminated in our patient cohort, we dichotomized patients based on the presence of significant proximal arm weakness, which we reasoned should be a defining feature of SPSMA, and infantile disease onset, which identifies patients with CDSMA. Our analysis found that the presence of significant proximal arm weakness did not predict mild or absent sensory abnormalities as would be expected for patients with SPSMA. Instead, patients reporting the greatest degree of shoulder weakness also reported more sensory symptoms in the hands and feet. Thus, we were not able to validate proximal arm weakness as a clinical sign to distinguish between CMT2C and SPSMA. These results suggest that CMT2C and SPSMA represent a broad and overlapping spectrum of non-congenital onset TRPV4-related neuromuscular disease with variable sensory involvement and variable distribution of motor weakness, but without reliably distinguishing clinical features. In contrast, our analysis of infantile onset patients suggests that congenital or infantile onset disease is associated with a distinct clinical phenotype that could be differentiated from later onset disease. Specifically, infantile onset patients reported less sensory loss and pain, less hand weakness, fewer voice abnormalities and a greater degree of difficulties with ambulation. They also showed minimal correlation of age and CMTES, suggesting the possibility of more limited disease progression in infantile-onset patients, at least with respect to specific elements assessed in the CMTES. However, it may be the case that more focused assessments of other aspects of disease, such as vocal fold weakness or proximal arm and leg weakness, would demonstrate progressive disease burden over time in infantile-onset patients. Normal developmental maturation in paediatric patients could also counterbalance progressive disease and give the false impression of stability. As such, outcome measures such as the CMTPedS that are scaled based on age-related normative values could be particularly informative. Notably, we found variability among age of onset with respect to specific mutations, with examples of both infantile and non-infantile onset for all mutations for which there was data for two or more patients. While the apparent difference in infantile and non-infantile onset patients require validation in prospective longitudinal studies, our results suggest that age of disease onset is likely to be an important consideration in clinical trial design for TRPV4-related neuromuscular disease.

Identification of clinical outcome measures and surrogate biomarkers

Given the potential for therapeutic TRPV4 inhibition as a treatment for TRPV4-related neuromuscular disease, identification of responsive clinical outcome measures is of critical importance. With respect to CMTES, we found a modest correlation between age and disease CMTES severity, but we also noted significant longitudinal variability of CMTES over time for individual patients. These findings may indicate challenges in applying a scoring system designed to detect length-dependent motor and sensory neuropathy to TRPV4-related neuromuscular disease, in which symptoms do not follow this pattern. As such, we favor evaluating alternative clinical outcome measures that are more sensitive to proximal arm and leg weakness and gait dysfunction. Specifically, the CMT functional outcome measure (CMT-FOM), CMTPedS and Overall Neuropathy Limitation Scale (ONLS) incorporate a broader range of lower extremity functional assessments and may better capture proximal weakness and related ambulatory difficulties.^35,48-52 With respect to the upper limbs, Medical Research Council (MRC) scores of proximal and distal muscles and clinical scales devised for SMA, such as the upper limb module, could be considered.⁵³ Based on accumulating evidence of the utility of MRI muscle fat fraction as a responsive outcome measure in multiple forms of CMT,^54,55 studies of muscle fat fraction in the lower limbs of TRPV4 patients should be explored.

Given the high prevalence and associated morbidity of vocal fold dysfunction and respiratory symptoms, these specific disease features warrant further study as potential clinical outcome measures. Whereas formal testing of respiratory function with spirometry could be readily employed in a longitudinal study, similar methods for quantitative assessments of vocal fold movement in TRPV4 patients will need to be identified. As vocal fold involvement in TRPV4 patients most commonly manifests as weakness of vocal fold abduction, the degree of vocal fold dysfunction could be amenable to quantification based on vocal fold excursion and maximal abduction angles.^56,57 Qualitative and quantitative speech and voice analysis could also be employed to assess speech impairment. Performing prospective and quantitative assessments of speech, vocal fold function and respiratory function will be important to determine whether these features are progressive over time, potentially amenable to treatment and ultimately useful as clinical outcome measures.

Limitations and future directions

While this work represents the largest cohort of TRPV4 neuromuscular patients analysed to date, there are important limitations to this study. First, there are likely to be ascertainment biases in our cohort due to increased motivation to participate in patients with severe disease and in parents of severely affected children. Such biases may have led to overestimation of the severity of disease and the proportion of early-onset patients. Notably, the patient data in our analysis is largely cross-sectional and thus provides a limited assessment of disease progression. While we were able to query patients regarding subjective worsening of various symptoms, such data was not verified by longitudinal neurological examinations. Our questionnaire data also relied in part on patient report of often complex and/or esoteric medical information, such as the presence of stridor, arthrogryposis and skeletal dysplasia, defining the age of disease onset, and differentiating muscle weakness affecting different parts of the limbs (e.g. shoulders versus hands). Thus, the validity of patient responses to such questionnaire items should be viewed with some degree of caution. Finally, longitudinal assessment of paediatric disease progression was severely limited in our cohort, as we had longitudinal CMTPedS scores for only 3 of 13 patients. Despite these potential shortcomings, our data provide a detailed multisystem analysis of a large TRPV4 patient cohort, providing insights into the prevalence of specific diseases manifestations, underscoring the importance of vocal fold dysfunction and skeletal disease and suggesting important differences among infantile and non-infantile onset patients. These collective findings will help guide ongoing prospective natural history and disease biomarker studies (NCT05600764) with the goal of designing a clinical treatment trial for TRPV4 patients.

Appendix 1

The Inherited Neuropathies Consortium (INC)

Eleonora Cavalca, Luca Crivellari, John Day, Matilde Laura, Stefania Magri, Isabella Moroni, Bipasha Mukherjee-Clavin, Emanuela Pagliano, Alex Rossor, Paola Saveri, Giulia Schirinzi, Mariola Skorupinska, Janet Sowden, Franco Taroni, Elizabeth Wood.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request. The data are not publicly available since they contain information that could compromise the privacy of research participants.

Funding

This work was supported by the National Institutes of Health (NIH) National Institute of Neurological Disorders and Stroke (NINDS) K08 NS102509 (B.A.M.), NIH NINDS R35 NS122306, (C.J.S.) and Muscular Dystrophy Association 629305 (C.J.S.). D.N.H., M.M.R., S.S.S. and M.E.S. were supported by the Inherited Neuropathies Consortium (U54 NS065712), a part of the NIH National Center for Advancing Translational Sciences (NCATS) Rare Diseases Clinical Research Network (RDCRN), an initiative of the Office of Rare Disease Research.

Competing interests

M.H., S.T., T.O.C., C.J.S and B.A.M. are participating in an investigator-initiated natural history study for TRPV4 neuropathy (NCT05600764) that is supported by Actio Biosciences. C.S. is supported by Actio Biosciences for laboratory-based studies of TRPV4 channelopathy. D.N.H. has served as a consultant for Regenacy, Applied Therapeutics, DTx Pharma, Passage Bio, Roche, Pfizer, Orthogonal Neurosciences, NMD Pharma, Sarepta, GLG, Guidepoint Global. D.P. acknowledges payments or reimbursements for consultancy and participation in Advisory Boards of Inflectis, Augustine Tx, and DTx Pharma and participation as site PI in clinical trial sponsored by AT Therapeutics.

Supplementary material

Supplementary material is available at Brain online.