Introduction

Erythromelalgia was first described in a 1878 paper by Silas Weir Mitchell entitled, “On a rare vasomotor neurosis of the extremities and on the maladies with which it may be confounded.“1 The term “erythromelalgia” was derived from the Greek words: “erythros” = red, “melos” = extremity, and “algos” = pain. In Mitchell’s observations, patient symptoms were primarily in the feet and characterized by a burning sensation that was alleviated by cold and exacerbated by warmth or physical activity, and associated with redness.

Since then, the diagnostic criteria for erythromelalgia have evolved. In 1932, Brown classified some cases as secondary to other underlying diseases and others as “primary”.2 He proposed four fundamental criteria for the diagnosis of erythromelalgia: “1) bilateral burning pain in the extremities, 2) sharp increase of local heat in the affected parts, but redness, flushing or congestion may vary in degree, 3) production and aggravation of the distress by heat and exercise, and 4) relief by rest, cold and elevation.”2 In 1938, Smith and Allen proposed substituting the term “erythromelalgia” for “erythermalgia” denoting the importance of the hot burning sensations.3 In 1994, Drenth and Michiels proposed three classifications: 1) erythromelalgia in thrombocythemia, 2) primary erythermalgia, and 3) secondary erythermalgia.4 More recently, the recognition that some cases of familial erythromelalgia are linked to dominant gain-of-function mutations of the SCN9A gene led to the introduction of the term “inherited erythromelalgia” referring to patients with a confirmed genetic variant, leaving “symptomatic erythromelalgia” as a descriptor for cases without a confirmed genetic variant.5

Today, erythromelalgia and erythermalgia are used interchangeably to describe a rare, chronic pain disorder characterized by the triad of intense burning sensation, warmth, and erythema, primarily involving the distal extremities (hands and feet), and usually alleviated by cold and worsened by heat.6,7 Patients with these findings commonly experience prolonged delay in diagnosis, suffer from missed diagnosis or remained undiagnosed. They are referred and evaluated across multiple subspecialties such as pain medicine, neurology, rheumatology, dermatology, and genetics.8 With an estimated incidence of 0.25-2 cases per 100,000 per year,9,10 it is a rare condition that is also associated with high morbidity. The rarity of the condition has led to few case series or cohort studies on pediatric erythromelalgia, and even fewer on their longitudinal trajectories.5,8,11 Overcoming these knowledge gaps in the etiology of erythromelalgia offers potential to develop targeted therapies to reduce pain and associated co-morbidities.

A comprehensive review encompassing all potential etiologies or presentations of pediatric erythromelalgia has not been published. General reviews have been written, but did not focus on clinical presentations in children.12,13 Therefore, the objective of this scoping review was to: 1) map the existing literature on erythromelalgia in children and adolescents, 2) identify knowledge gaps, and 3) inform directions for future research in pediatric erythromelalgia.

Methods

The scoping review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews,14 and pre-registered on Open Science Framework.15

We searched five online databases: PubMed, Embase, Web of Science, CINAHL, and Cochrane. The selection of articles was made through the following search string: (“erythromelalgia” OR “erythermalgia”). Articles published from inception until November 30, 2023 were extracted on December 11, 2023.

Two reviewers independently screened the titles and abstracts of all extracted articles matching the research aim and inclusion criteria, excluding duplicates, reviews, commentaries, posters, and proceeding papers, using Covidence (www.covidence.org). Original English pediatric (< 18 years old), single case studies, case series (i.e. ≥ 2 cases presented), and cohort studies, or adult single case studies, case series, and cohort studies with symptoms that emerged at the age of 18 years old or younger were identified for full-text screening. The reference list of all screened reviews and full-text articles were checked for additional papers matching the inclusion criteria. References deemed relevant by at least one reviewer underwent full-text screening. Any disagreements were resolved by a third independent reviewer.

Study characteristics (e.g. year of publication, country of authors, type of study) and all available data on demographics, clinical features, genetic and laboratory testing results, comorbidities, and management were extracted, entered, and cross-checked in a database on Covidence. Unreported data were labeled “not reported”. All data are presented as the frequency of studies (S) or cases (C), unless otherwise specified.

Results

Study selection

A total of 3,081 references were identified through database searching (Supplementary Fig. 1A). After removal of duplicates, 1,848 titles or abstracts were screened, leading to 732 articles that underwent full-text review. Eighteen additional articles, identified through the reference lists of screened reviews and articles, underwent full-text review, and were assessed for eligibility. There were 193 articles retained for data extraction. Pediatric-specific data could not be extracted in 19 articles, but their overall findings are summarized in Supplementary Table 1. Overlap in case reports led to 167 studies retained for analysis (Supplementary Table 2).

Study characteristics

The studies identified were primarily conducted in North America (S = 59), Europe (S = 53), and Asia (S = 34), followed by South America (S = 4), and Australia (S = 4). There were 11 multi-national studies as follows: United States and China (S = 4), Japan (S = 2) or Taiwan (S = 1), Netherlands and Belgium (S = 1) or Canada (S = 1), United Kingdom, United States and Germany (S = 1), and Germany, Norway and Sweden (S = 1). The affiliation of the authors of one study was not reported. The types of studies published included case reports (S = 100), case series (S = 41), translational studies (S = 22), case-control studies (S = 2), and randomized, double-blind, placebo-controlled, crossover trials (S = 2). Although the first confirmed case of pediatric erythromelalgia was published in 1878, most of the studies identified (S = 125) reporting pediatric cases of erythromelalgia were published in the last two decades (Supplementary Fig. 1B).

Childhood-onset erythromelalgia were met for 411 cases reported in these studies, in which 148 were male, 229 were female, and 34 were not reported. Race and ethnicity were not reported for most cases (C = 355), but the race of those reported included Asian (n = 19), Black or African American (n = 1), Hispanic or Latino (n = 1), and White (n = 35). Familial inherited erythromelalgia was reported in 51 studies, while spontaneous or de novo erythromelalgia was reported in 48 studies. Family history was not reported in 68 studies. Overall, the studies comprise 189 cases with primary inherited erythromelalgia (confirmed family history or genetic mutation), 194 with primary symptomatic erythromelalgia, and 28 with secondary erythromelalgia (6 Olmsted syndrome, 7 small fiber neuropathy, 3 small fiber neuropathy and Familial Mediterranean fever, 1 small fiber neuropathy and Behcet’s disease, 6 acute secondary erythromelalgia, 1 red ear syndrome – erythromelalgia type, 2 erythrocyanosis, 1 painful redness of feet due to chilblains, and 1 DiGeorge and CHARGE syndromes).

Outcomes reported

Clinical presentation

Erythromelalgia was usually diagnosed based on clinical history and physical examination. The diagnostic approach was primarily focused on excluding other underlying diseases, and some studies also included genetic testing. Overall, the main criteria for diagnosis of erythromelalgia followed Brown’s 1932 criteria which includes: 1) burning pain of the extremities, 2) pain aggravated by warming and exercise, 3) pain relieved by cooling, rest or elevation, 4) redness of the affected skin, and 5) increased temperature of the affected skin. However, only 53 studies reported all five diagnostic criteria in their cases (Table 1) representing 163 of all cases reported. Relevant physical examination findings were also documented for many patients: swelling/edema was reported in 69 studies (C = 108), and skin injury (ulcers, lesions, blisters, maceration, erosion, and cracking) was reported in 59 studies (C = 152).

Table 1 Characteristics of cases reported in studies.
Full size table

The studies reported cases mainly affected in their hands (C = 236), feet (C = 358), and ears (C = 45), or hands and feet (C = 230) (Table 1). Other locations included the neck, trunk and groin. Most of the studies did not report any inciting event (S = 130) or reported the spontaneous onset of erythromelalgia symptoms (C = 10). Reported inciting events (C = 62) prior to the erythromelalgia symptoms included illness (C = 27), infection (C = 33), physical activity (C = 23), trauma (C = 6), surgery (C = 1), post-vaccination (C = 2), ingestion of Agaricus spp. and Lepista inversa (Scop.) Pat mushrooms (C = 1), and cessation of norephedrine therapy (C = 1). The medical history was not reported in 276 cases and noted as unremarkable in 55 cases. The remaining 80 cases with reported comorbidities (Table 1) included inflammatory (C = 18), neurologic (C = 17), vascular (C = 27), or other conditions (C = 40).

Psychosocial factors were assessed in 23 studies, but findings were reported in only 21 studies, which primarily included anxiety, behavioral problems, depression, and suicidal ideation (Table 1). Poor school/work attendance or performance was reported in 31 studies (C = 65). Impaired quality of life was reported in 107 studies (C = 211), which included difficulty sleeping, limited physical activity, limited social activity, lifestyle change (e.g. wheelchair-bound, loss of autonomy, moving environment), and preference to walk barefoot (Table 1).

Evaluation of genetic mutations

From this search, the first pediatric case of inherited erythromelalgia linked to a dominant gain-of-function mutation of the SCN9A gene was reported in 2004.16 Since then, only 70 studies have reported genetic testing, with 58 studies reporting a genetic candidate linked to the clinical profile of the cases. In 144 cases with childhood-onset erythromelalgia, 32 SCN9A gene mutation variants were reported, while 9 non-SCN9A genetic mutations were reported for 11 cases (Table 2).

Table 2 Genetic mutations reported in pediatric cases.
Full size table

Clinical investigations

Laboratory tests were reported in 106 studies (Table 3). Abnormal laboratory findings included: high platelet count, increased erythrocyte sedimentation rate and/or C-reactive protein, positive antinuclear antibodies, high white blood cells, defect in immunity (e.g. low levels of immunoglobulins or complement deficiency), and elevated liver enzymes (Supplementary Table 3). Neurological examinations were reported in 110 studies (Table 3). Significant findings included: abnormal quantitative sensory testing results, sudomotor dysfunction, abnormal nerve conduction, small fiber neuropathy, hypertension (i.e. high resting blood pressure), and low intraepidermal nerve fiber density in skin biopsies (Supplementary Table 3). Other significant non-neurological findings from skin biopsies included inflammation of the blood vessels (e.g. leukocytoclastic vasculitis, strong phospho-extracellular signal-regulated kinase expression, perivascular or interstitial infiltrate, thick capillary walls) and hyperkeratosis. Vascular studies were reported in 55 studies (Table 3). Abnormal vascular findings included: increased blood flow and associated increased temperature, decreased blood flow, and abnormal morphology (Supplementary Table 3). Imaging studies were reported in 52 studies (Table 3). Abnormal imaging findings included: bone loss (C = 2), scans suggestive of reflex sympathetic dystrophy or complex regional pain syndrome (C = 2, e.g. bone marrow edema and increased tracer accumulation in the affected areas at the vascular, blood pool and bone phases), spina bifida occulta (C = 1), structural or functional changes in the head/neck at rest (C = 5, e.g. focal epileptiform discharges with diffuse background slowing, cervical spondylopathy, asymmetry in the anterior horns of the ventricles, abnormalities of the pituitary stalk, a small adenohypophysis, reduced depth of the sella turcica, or cerebral infarctions) or during treatment/pain relief (C = 2, e.g. changes in cerebral blood flow between baseline pain and cooling relief, or shift in brain activity during carbamazepine treatment from valuation (i.e. decision-making) and pain areas toward primary somatosensory-motor and parietal attention areas).

Table 3 Clinical investigations reported in studies.
Full size table

Treatment approaches

Various combinations of pharmacological and non-pharmacological treatments and related responses were reported across most studies. There were 97 cases of which there was symptom resolution (i.e. complete “relief” or absence of symptoms related to erythromelalgia including pain, redness, edema, and heat, as well as return to normal activities) through procedural interventions with systemic pharmacotherapy (C = 3), procedural interventions (C = 21), psychological approaches (C = 1), and pharmacotherapy (C = 72). Of these resolved cases, only 18 cases had a confirmed genetic mutation (2 TRPV3 mutations and 16 SCN9A mutations). If follow-up was noted, “symptom resolution” was reported 1 week up to 5 years post-treatment. Procedural interventions were reported in 56 studies (Table 4) which primarily included: neural axial blockade (i.e. epidural catheters insertion), infusions, nerve blocks, sympathectomies, and transcutaneous electrical nerve stimulations. Non-pharmacological approaches included avoidance of triggers (e.g. reduce exposure to heat), physiotherapy, psychology, pain rehabilitation programs, and other treatments (Table 4). Pharmacotherapy was reported in 148 studies (Table 4 and Supplementary Table 4). The most common pharmacological classes included adrenergic agonists, antihistamines, beta-blockers, calcium channel blockers, corticosteroids, cyclooxygenase (COX) inhibitors, opioid receptor agonists, sodium channel blocker, and selective serotonin/serotonin-norepinephrine reuptake inhibitors (SSRIs/SNRIs).

Table 4 Therapeutic approaches reported in studies.
Full size table

Discussion

To our knowledge, this is the first scoping review to map the existing literature in pediatric erythromelalgia. Our goal was to identify gaps in knowledge to inform future research. A total of 167 studies totalling 411 individuals with childhood-onset erythromelalgia were identified. Findings suggest contrasting clinical presentations, assessment, and treatment of pediatric erythromelalgia.

Only 53 studies reported on all five of Brown’s 1932 diagnostic criteria in their cases, representing 40% of the cases.2 This illustrates the significant variability in clinical presentation of pediatric erythromelalgia cases, highlighting its diagnostic challenge. Case-control studies and clinical trials identified were not usually specific in their inclusion criteria17,18,19 such as “documented diagnosis of erythromelalgia, as characterized by redness, warmth, and burning pain of the extremities (most commonly feet), typically precipitated by heat or exercise and relieved by cooling.”20 Moreover, cases have reported their main affected areas as the face, ears or groin, which are not usually included in the definition of extremities, or have reported their symptoms as episodic or continuous. Future directions should test the sensitivity and specificity of diagnostic criteria of EM, and whether there are differences for those who present symptoms during the pediatric lifespan. We propose at least three of five criteria based on Brown’s diagnostic criteria must be met for a presumptive diagnosis of erythromelalgia (episodic or continuous) (Fig. 1). Diagnosing erythromelalgia in youth is a critical step towards its recognition and validation, and may lead to its treatment and the opportunity for the patient to connect with support groups (e.g. The Erythromelagia Association, The Erythromelalgia Warriors, Ben’s Friends - Living with Erythromelalgia online community, etc.) and identify self-management strategies.

Fig. 1: Proposal for diagnostic workup and classification for pediatric erythromelalgia and treatment approach.
figure 1

Burning pain, redness, and increased temperature of the affected skin can be either continuous or episodic.

Full size image

Contrasting hypotheses on the pathophysiology of pediatric erythromelalgia have led to widely varying treatment approaches. Tham et al. conducted a critical review of current pain management approaches for patients with erythromelalgia and concluded there were no established best practices or clinical guidelines for pain management in this disorder.7 Similar to approaches identified in our scoping review, management included avoiding situations that may precipitate pain, as well as incorporating pharmacotherapy, procedural interventions, and non-pharmacological interventions.7,12,13 Qualitatively, findings from our scoping review highlight that the current treatment of pediatric erythromelalgia is a stepwise trial and error approach. Ma et al. published a proposed approach to management including non-pharmacological and pharmacological (topical and systemic) strategies.21 They highlight that labelling erythromelalgia as secondary or primary would not affect the treatment approach significantly. However, several specific subtypes have been identified for which making a mechanistic diagnosis can lead to more targeted treatment.18,19,20,22,23,24,25 Based on these reports, we propose a modified approach for youth with erythromelalgia (Fig. 1).

Upon evaluation, it is important to treat or manage any underlying causes or associations (e.g. myeloproliferative diseases). Laboratory tests and imaging studies for most cases reported in this scoping review were primarily used to exclude secondary diseases associated with erythromelalgia. The scoping review highlights the preponderance of negative findings from laboratory tests and imaging studies in youth with erythromelalgia. As an example, testing for Fabry disease, where supported by features of the history and exam, has therapeutic importance because of the availability of a specific enzyme replacement therapy.26,27 However, among the cases and unpublished experience of the co-authors, Fabry disease is likely to account for only a very small fraction of patients presenting with erythromelalgia. Neurological and vascular examination findings were consistent with literature in adult populations. There was a small proportion with abnormal findings of increased blood flow and temperature, or large and/or small fiber neuropathy.28,29 The diverse and limited positive findings highlight the overall heterogeneity in clinical presentation of erythromelalgia which has led to contrasting views of its underlying pathophysiology (neurologic, inflammatory, autoimmune, vascular, etc.). Erythromelalgia associated with thrombocythemia has been linked to gene variants producing a constitutively activated form of Janus Kinase 2 (JAK2).22 Essential thrombocythemia is one of many myeloproliferative disorders whose symptoms may include red, warm, burning, or tingling hands or feet. Patients with erythromelalgia with associated thrombocythemia have displayed relief by aspirin due to its irreversible inhibition of platelet cyclooxygenase activity.22 However, the lack of longitudinal trajectories of youth with erythromelalgia and low prevalence of high platelet counts from available patient samples argues against the hypothesis that thrombocythemia is a frequent cause of erythromelalgia.30

Oaklander has postulated that erythromelalgia may be driven primarily through small-fiber neuropathy associated with systemic autoimmune/autoinflammatory disorders.31 Conventional electrodiagnostic tests such as electromyography or nerve conduction studies are insensitive to small-fiber neuropathy. Skin biopsies are safe and minimally invasive, and have been recommended by Oaklander and others as useful in diagnosis of small-fiber neuropathy. Case reports of erythromelalgia with small-fiber neuropathy have reported improvements after intravenous immunoglobulin therapy or corticosteroids.32,33 However, there are limited reference values for skin neurite densities in healthy children, and there are uncertainties about how to interpret positive and negative predictive values.34,35,36 A randomized controlled trial in 30 adults with painful idiopathic small fiber neuropathy found that intravenous immunoglobulin treatment had no significant effect on pain compared to placebo control.37 Our scoping review highlights that clinicians have limited information to guide them on how best to use clinical variables to prioritize laboratory testing for other diseases that could present as erythromelalgia. Future consensus is warranted to determine the diagnostic procedure for youth displaying the signs and symptoms suggestive of erythromelalgia.

In 2004, several cases of Mendelian inherited erythromelalgia were linked to autosomal dominant gain-of-function mutations of the SCN9A gene. SCN9A encodes for the voltage-gated sodium channel Nav1.7, which is found primarily in small peripheral sensory and sympathetic neurons.16 Gain-of-function mutations in SCN9A lead to an increase in cellular sodium influx resulting in increased signaling in nociceptive neurons. In a recent systematic review, Arthur et al. identified 16 different substitutions of Nav1.7 channels,5 while up to 32 SCN9A gene variants have now been identified in 146 cases in our scoping review. The increase in novel SCN9A gene variants can be attributed to an increase in recognition/publication of erythromelalgia, advancement in genetic analyses, and growth of genetic databases. For patients with SCN9A variants, studies suggest that the electrophysiological properties of the specific variant channel may predict responsiveness to sodium channel blockers such as carbamazepine, oxcarbazepine, and mexiletine.23,24,25 However, in our scoping review, mutations of the SCN9A gene were identified as the cause of only 35% of reported pediatric cases of erythromelalgia. Therefore, the majority of cases may have non-genetic etiologies or a yet unidentified mutation in one or more genes. Plausible candidates could be genes involved upstream or downstream of Nav1.7 activity or previously implicated in neurological, inflammatory, vascular, or pain disorders,38,39,40 such as Familial Episodic Pain Syndrome (TRPA1).41 Adult erythromelalgia case reports have identified genetic variants that alter the function of other voltage-gated sodium channels (SCN10A and SCN11A),42 or platelet-endothelial interactions (e.g. JAK2),22 as noted above. The identification of novel variants related to erythromelalgia will have broader implications in understanding pain mechanisms in general and may hopefully lead to novel approaches to pain treatment, as previously done with Nav1.7.43

Based on the findings of this scoping review, the levels of evidence for treatment for youth with erythromelalgia are considered low. Topical treatments are usually considered as a first-line pharmacological treatment, since they cause fewer adverse effects compared to systemic medications and interventional procedures.21 In our scoping review, medications used for neuropathic pain, including antidepressants, anticonvulsants, and sodium channel blockers were commonly prescribed for youth with erythromelalgia. Evidence for these drug classes for neuropathic pain in children is sparse,44 and prescribing is largely based on extrapolation from adult studies of other forms of neuropathic pain. While gabapentin is often the first medication prescribed for neuropathic pain in other settings, it is the practice of several of the authors of this review to select a sodium channel blocker, e.g. oxcarbazepine or mexiletine, as the first neuropathic medication for a trial in children with erythromelalgia.

When underlying causes or associations cannot be determined, it is important to provide counselling and non-pharmacological strategies addressing psychosocial factors involved with erythromelalgia. Psychosocial factors and quality of life were reported in only 65 and 211 cases, respectively. Recognizing that pain is a complex multidimensional experience that is the result of interactions between biological, psychological and social factors,45 highlights the need to study the perspective and experiences of youth with erythromelalgia. Our scoping review highlights significant co-morbidities in youth with erythromelalgia which include anxiety, depression, suicidal ideation, and sleep impairment, alongside physical and social limitations. Despite the small proportion of cases reporting mental health comorbidities, studies focusing on its co-occurrence with pediatric chronic pain have shown higher mental health issues in youth with chronic pain compared to their pain-free peers.46 Recognizing the prevalence of these factors and addressing them with a multidisciplinary team that includes mental health providers is essential for these cases. Therefore, future research should incorporate core domains and outcome measures47 for pediatric chronic pain trials and registries that encompass measures of pain severity, pain-related interference, overall well-being, emotional and physical functioning, and sleep.48,49

Referral to a comprehensive multidisciplinary pain rehabilitation center could be considered for youth with severe, refractory, or disabling EM. However, whether exercise exacerbates the symptoms of erythromelalgia is important to consider for rehabilitation, especially as this was reported for nearly 60% of the cases in this scoping review. Nevertheless, understanding the pathophysiology of pediatric erythromelalgia and identifying clinical subgroups within a large sample of pediatric erythromelalgia patients may offer an initial step to determining individualized therapeutic approaches.

Most of the reports in this scoping review were retrospective in nature. This has inherent limitations, as the data are primarily drawn from text fields with descriptive or narrative responses, and there is missing data. Structured and/or standardized inputs for diagnostic criteria will be beneficial for future research purposes. As pediatric erythromelalgia is a rare condition, the development of an international registry would immensely benefit multidisciplinary experts involved in the care of pediatric erythromelalgia and those with lived experience. In this regard, our team is developing a multicenter PEDiatric ErythoMElalgia Registry Gathering multidisciplinary Experts (PED-EMERGE) to investigate our hypothesis that erythromelalgia is a clinical syndrome that includes multiple mechanisms in distinct patient subgroups. Moreover, our team recruits pediatric erythromelalgia patients to undergo genetic screening which may lead to the discovery of new Mendelian causes of erythromelalgia or predisposing genes that are conserved across pediatric cases. Our team includes patient partners involved in ensuring research projects and the registry are patient-centered. The objective of this consortium is to create collaborations between diverse experts and generate patient-centered clinical effectiveness research projects.

This scoping review revealed variability in the clinical presentation of pediatric erythromelalgia regarding diagnostic criteria, clinical examination findings and treatments offered. Ongoing efforts focus on developing a multicenter registry to standardize data collection and reporting with the goal of establishing consensus recommendations for the diagnosis and management of pediatric erythromelalgia.