Childhood-onset hypocomplementaemic urticarial vasculitis in France: phenotypic and genotypic diversity in 10 children

Objectives

Hypocomplementaemic urticarial vasculitis (HUV) is a rare small-vessel vasculitis characterized by the combination of urticarial vasculitis and hypocomplementaemia. HUV may occur in isolation or be associated with systemic diseases such as systemic lupus erythematosus (SLE). Recently, loss-of-function variants in DNASE1L3 have been reported to cause familial forms of HUV. We aimed to describe the clinical and genetic features of childhood-onset HUV cases in France.

Methods

We conducted a nationwide retrospective study in France that included 10 patients with childhood-onset HUV.

Results

Ten children, all girls, were included. Their median age at onset was 9.5 years (range: 2–14 years), and the median delay between onset and diagnosis was 13 months (range: 6–55 months). Phenotypes were heterogeneous: 3 had systemic HUV syndrome (S-HUVS), all associated with DNASE1L3 deficiency, including 2 siblings; 4 had non-systemic HUVS (NS-HUVS), one of whom had a C2 deficiency; and 3 had SLE-associated HUV. Musculoskeletal features were the most common extracutaneous manifestation (9/10). The most severe manifestations were observed in children with S-HUVS, including pulmonary haemorrhage (2/3) and glomerulonephritis (2/3). Anti-nuclear antibodies were positive in 9/10 patients, anti-C1q antibodies in 6/10, and anti-neutrophil cytoplasmic antibodies in 5/10. Antihistamines were widely used in NS-HUVS, while S-HUVS and SLE-associated HUV required immunosuppressive treatments. Hydroxychloroquine was used in 8/10 patients.

Conclusion

These 10 cases highlight the heterogeneity and potential severity of childhood-onset HUV. We report the third familial form of HUVS associated with DNASE1L3 deficiency and an extremely rare case of HUVS associated with C2 deficiency.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12969-026-01203-w.

Key messages

Childhood-onset HUV is a rare disease with phenotypic and genotypic heterogeneity.

Systemic HUVS, associated with DNASE1L3 deficiency but not with positive anti-C1q, is the most severe form.

A genetic testing should be systematically performed in childhood-onset HUV.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12969-026-01203-w.

Introduction

Urticarial vasculitis (UV) is an immune complex small-vessel vasculitis characterised by distinctive skin lesions that differ from common urticaria. In UV, urticaria is fixed, persists for more than 24 h, and resolves with ecchymosis post-inflammatory hyperpigmentation [1]. Hypocomplementaemic UV (HUV), also known as anti-C1q vasculitis according to the 2012 revised International Chapel Hill Consensus Conference nomenclature of vasculitis [2], is defined as the coexistence of UV and hypocomplementaemia. McDuffie first described HUV in 1973 as a persistent autoimmune disorder characterised by cutaneous vasculitis, hypocomplementaemia and systemic manifestations [3].

In both adults and children, HUV can occur in association with systemic diseases, such as systemic lupus erythematosus (SLE) or as an isolated condition, referred to as HUV syndrome (HUVS) or McDuffie syndrome [4, 5]. Schwartz et al. proposed two essential criteria for diagnosing HUV: the presence of typical urticarial skin lesions lasting more than 6 months and reduced serum complement levels. Additionally, at least two of the following supplementary criteria are required for diagnosis: leukocytoclastic vasculitis on skin biopsy, arthralgia and/or arthritis, glomerulonephritis, ocular inflammation, recurrent abdominal pain, and the detection of anti-C1q antibodies [6].

The physiopathology of HUV remains poorly understood and implies complex immunological mechanisms. The deposition of immune complexes in blood vessels is thought to trigger inflammatory responses, leading to vasculitis [7]. Immune complexes also contribute to overactivation of the classical complement pathway, causing its rapid consumption. In particular, C1q levels are reduced in patients with HUV due to this complement activation [1, 4]. Recent research has identified genetic factors, including loss-of-function variants in the DNASE1L3 gene (NM_004944.3), in familial forms of HUVS in 2013 [8]. These variants, previously linked with familial forms of SLE in 2011, abrogate the activity of the endonuclease DNASE1L3, which plays a role in disposing of DNA from dying cells [8–10]. The DNA that accumulates in the absence of DNASE1L3 leads to the development of anti-DNA autoreactivity [10, 11].

HUV is a rare disease. Jachiet et al. reported the largest series of adults with HUV including 57 patients [4]. In children, the analysis of reported cases is complex, as case series are generally not exhaustive and include either only children with SLE or only those with DNASE1L3 deficiency. Moreover, these different series overlap by including children already described in others. Al Musalhi et al. documented the largest series of children (n = 54) but all had HUV associated with SLE, and they included the first 17 patients with DNASE1L3 loss-of-function variants previously described by Al-Mayouf et al. [5, 9]. Tusseau et al. described five new children with DNASE1L3 deficiency and conducted a global review of the 35 previously reported cases of DNASE1L3 deficiency, raising the total number of children identified with DNASE1L3 deficiency associated with SLE and/or HUVS to 40 [12]. Sometime later, Abdwani al. published a series of 33 children with SLE associated with DNASE1L3 deficiency from 15 families [13], most of whom had already been described elsewhere [5, 9], including in the series by Tusseau et al. [12]. More recently, Taş et al. focused their review on paediatric cases of HUVS +/- associated with SLE, and +/- associated with DNASE1L3 deficiency, raising the total to 28 children previously described with HUVS, 2 of whom died [14]. These case studies highlight the scarcity, heterogeneity and potential severity of HUV in children.

In this study, we described 10 cases of childhood-onset HUV in France, with a specific focus on two new siblings exhibiting DNASE1L3 deficiency, and one child with C2 deficiency.

Methods

Patients

All physicians affiliated with the French-speaking Society of Rheumatology and Paediatric Inflammatory Diseases (SOFREMIP) were contacted, using the association’s mailing list, to collect cases of childhood-onset HUV. The heads of paediatric rheumatology departments at the largest hospitals were also contacted directly. The response rate was 100%.

Patients were retrospectively included if they met the criteria proposed by Schwartz et al. [6].

Collected data

Demographic, clinical and laboratory data, as well as treatments and disease outcomes, were retrospectively collected. The following clinical data were recorded at diagnosis and during follow-up: sex, age at disease onset, age at diagnosis, characteristics of cutaneous involvement, presence of systemic features and organ involvement. The following biological data were also collected at diagnosis and during follow-up: levels of complement fractions (C2, C3, C4, CH50 and C1q), presence of anti-C1q antibodies, anti-nuclear antibodies (ANA), anti-extractable nuclear antigen antibodies, anti-double-stranded DNA antibodies (anti-dsDNA), anti-neutrophil cytoplasmic antibodies (ANCA) and genetic variants identified as pathogenic.

Definitions

Manifestations, such as urticaria and organ involvement, were defined as previously described [6]. Systemic features were defined as fever, weight loss and/or asthenia. Biological parameters were determined according to the standards of the laboratory that performed the analysis. Hypocomplementaemia was defined as a C3 and/or C4 fraction below the laboratory reference range, at the time of diagnosis and on at least one other occasion. The 2012 Systemic Lupus International Collaborating Clinics criteria were used to confirm the diagnosis of SLE [15]. The two HUV subgroups were defined as follows: HUVS, when isolated, and SLE-associated HUV (SLE-HUV), when HUV was associated with SLE. HUVS cases were further divided into two subgroups: non-systemic HUVS (NS-HUVS) and systemic HUVS (S-HUVS). S-HUVS was defined by the involvement of at least two organs, including ocular, pulmonary, gastrointestinal or renal involvement [4].

Genetic analysis

Written informed consent for the genetic analyses was obtained from the parents. DNA was extracted from peripheral blood samples. Sanger sequencing was performed in patient 2 (P2) and P3 on all coding exons of DNASE1L3 to assess for pathogenic variants. The EF Library Prep enrichment kit (Twist Biosciences, South San Francisco, CA, USA) was used, along with either the Human Comprehensive Exome (P6, P10) (Twist Biosciences) or a custom probe panel (P1, P5, P8) (Twist Bioscience, Agilent) to capture regions of interest in accordance with the manufacturer’s instructions. Sequencing was performed on a NextSeq 500 System (Illumina, San Diego, CA, USA). Bioinformatic analysis was conducted according to the Broad Institute Genome Analysis Toolkit best practice guidelines [16]. The pipeline for variant calling and prioritisation has been described elsewhere [17]. Briefly, priority was given to rare exonic or donor/acceptor splicing variants consistent with the pedigree and phenotype and/or variants in genes previously associated with autoinflammatory phenotypes. Variants were classified according to the recommendations of the American College of Medical Genetics and Genomics [18]. cDNA and protein sequence variants were described following the recommendations of the Human Genome Variation Society, using NCBI reference sequences and genomic coordinates from GRCh37/hg19. Variants identified through next-generation sequencing, as well as family member segregation, were confirmed by Sanger sequencing.

DNASE1L3 activity assay

Purified nuclei from MCF7 breast cancer cell line (2.10⁵) were left untreated or treated with recombinant DNASE1 (5 and 50ng), DNASE1L3 (5 and 50ng) or human plasma (diluted 1/6) in presence of CaCl2 and MgCl2 (2mM) for 2 h at 37 °C. Upon digestion DNA was purified using QIAamp DNA Blood Mini Kit (Qiagen), and nuclear DNA fragmentation was evaluated by agarose (2%) gel electrophoresis.

Human ethics approval and consent to participate

This retrospective study was conducted under the French data protection authority MR004 reference methodology, was approved by the Ethical Committee of the Bordeaux university hospital centre (CER-BDX 2025 − 240), and was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Parents and patients provided their written informed consent to participate.

Statistical analysis

Descriptive statistics included median (minimum–maximum) for continuous variables.

Results

Patients

Ten children, all female, were included. Median age at onset was 9.5 years (2–14). Median age at diagnosis was 10 years (4–16), with a median delay of 13 months (6–55) between onset and diagnosis. The median follow-up period after diagnosis was 26.5 months (18–87). All 10 patients presented with chronic urticarial skin lesions, which were pruritic in 8 patients. Urticaria was associated with angioedema in 9 patients, and with purpuric lesions in 5 patients. Skin biopsies were performed in 5 patients, revealing leukocytoclastic vasculitis in 3 patients, and neutrophilic dermatosis in the other 2 patients. Arthralgia and/or arthritis were observed in 9 patients, glomerulonephritis in 3 patients, recurrent abdominal pain in 3 patients and ocular inflammation in 2 patients. All patients had low levels of C3 and/or C4, 7 had low C1q, and 6 tested positive for anti-C1q antibodies. ANA were positive in 9 patients, and ANCA in 5 patients. Based on these characteristics, 3 patients (P1–P3) were classified as having S-HUVS, 4 patients (P4–P7) as having NS-HUVS, and 3 patients (P8–P10) as having SLE-HUV. Their characteristics are presented in Table 1, and detailed in Supplemental Table S1.

Table 1.

Characteristics of 10 children with HUV

Cells are coloured to indicate the presence of the characteristic in the patient: red represents S-HUVS, yellow represents NS-HUVS and blue represents SLE-HUV

HUV, hypocomplementaemic urticarial vasculitis; S-HUV, systemic HUV; NS-HUV, non-systemic HUV; ANCA, anti-neutrophil cytoplasmic antibodies; dsDNA, double-stranded DNA; NI, not investigated; F, female

Systemic HUVS

P1, P2 and P3 were 2, 5 and 10 years old at disease onset, and 4, 6 and 11 years old at diagnosis, respectively. The mean delay between onset and diagnosis was 16 ± 9 months.

P1 and P2 were siblings with HUVS associated with a DNASE1L3 loss-of-function variant. Their healthy parents, originally from Iraq, were not known to be consanguineous but came from the same village. P1 presented with recurrent fever, urticarial-like lesions, angioedema, and conjunctival hyperaemia from the age of 2 (Fig. 1). A skin biopsy revealed perivascular and interstitial infiltration of neutrophils. At the age of 4, she developed severe intra-alveolar haemorrhage and glomerular vasculitis without complement deposits (Fig. 2). Laboratory investigations showed an ANA titre of 1:160, negative anti-dsDNA, an ANCA titre of 1:320 (positive anti-myeloperoxidase), and low levels of C3, C4 and C1q. Anti-C1q antibodies were not initially investigated but were subsequently investigated three times during flare-ups of the disease, and were always negative. Ophthalmological examination revealed conjunctival hyperaemia and significant papilledema with normal visual acuity. Brain magnetic resonance imaging and lumbar puncture results were normal. P2 presented with recurrent fever and chronic urticarial-like lesions starting at the age of 5. Eight months later, she developed profound anaemia, intra-alveolar haemorrhage, bilateral conjunctival hyperaemia associated with significant papilledema, and arthritis. Laboratory findings included an ANA titre of 1:320, negative anti-dsDNA and ANCA, low levels of C4 and C1q, and negative anti-C1q antibodies. Brain magnetic resonance imaging and lumbar puncture results were normal. Genetic analysis identified a homozygous DNASE1L3 (c.290_291del) variant previously reported as pathogenic in both siblings [8]. P1 and P2 exhibited no specific DNASE1L3 enzymatic activity in their circulation, as measured by the ability of plasma to digest nuclear DNA (Fig. 3). Familial segregation confirmed that both parents were heterozygous for this variant. Treatment with steroids, hydroxychloroquine and rituximab achieved partial remission in both siblings. After more than 2 years of follow-up, the intra-alveolar haemorrhage, renal involvement, papilledema, arthritis and recurrent fever resolved, but the skin involvement persisted in both siblings.

Fig. 1.

Skin involvement in hypocomplementaemic urticarial vasculitis syndrome. Angioedema and erythematous lesions on the foot (left) and on the abdomen (right)

Fig. 2.

Hypocomplementaemic urticarial vasculitis syndrome complicated by intra-alveolar haemorrhage. Reticular opacities on chest X-ray (left) and a diffuse tree-like appearance with buds on chest computed tomography (right)

Fig. 3.

Patient 1 (P1) and P2 with homozygous DNASE1L3 (c.290_291del) variant do not exhibit circulatory DNASE1L3 activity. Gel electrophoresis of DNA originating from purified nuclei that were left untreated, placed in presence of recombinant DNASE1L3 (D1L3), DNASE1 (D1), and human plasma from healthy donors (HD) and HUVS patients (P1 and P2)

P3 was a child of consanguineous parents (first cousins) of Arab origin. She was one of the five new patients (Patient 2) described by Tusseau et al. [12]. She presented with chronic urticaria, angioedema and arthralgia from the age of 10. Seven months later, she developed renal involvement. Laboratory investigations showed an ANA titre of 1:320, positive anti-dsDNA, an ANCA titre of 1:160 (positive anti-myeloperoxidase), low levels of C3 and C4 and absence of anti-C1q antibodies. A skin biopsy revealed leukocytoclastic vasculitis with immune deposits, and a renal biopsy showed proliferative extra capillary glomerulonephritis with immune deposits. Treatment with steroids, mycophenolate mofetil, tacrolimus and hydroxychloroquine achieved remission after more than 3 years of follow-up. Genetic analysis identified a homozygous DNASE1L3 c.433G > A variant, which was considered pathogenic [12]. Familial segregation confirmed that both parents were heterozygous for this variant.

Non-systemic HUVS

P4, P5, P6 and P7 had NS-HUVS. They were 13, 5, 9 and 11 years old at disease onset and 16, 9, 9 and 12 years old at diagnosis, respectively. The mean delay between onset and diagnosis was 28 ± 19 months. All patients presented with recurrent urticaria and angioedema, and 3 also had recurrent purpuric lesions. Skin biopsies, performed in 2 patients, revealed leukocytoclastic vasculitis in both cases. Other manifestations included arthralgia and/or arthritis (3/4) and recurrent abdominal pain (2/4). Laboratory investigations revealed isolated positive ANA in P4, P5 and P6, with titres of 1:320, 1:1280 and 1:640, respectively. Anti-dsDNA antibodies were negative in all patients. Only P5 had a positive ANCA titre (1:160, negative anti-myeloperoxidase and anti-proteinase 3). C1q levels were low, and anti-C1q antibodies were positive in 3 patients.

C2 deficiency was identified in P5. Both parents were healthy. The disease had begun at the age of 5 with recurrent UV lesions, purpura, abdominal pain, and arthralgia. No recurrent bacterial infections were reported. Laboratory investigations revealed positive ANA, ANCA and anti-C1q antibodies. Next-generation sequencing identified compound heterozygous variants in C2 (c.841_849 + 19del/c.1902 + 5G > A), with each variant inherited from one parent. The c.841_849 + 19del variant, commonly referred to as Del28pb, has previously been reported as pathogenic, causing loss of function due to premature protein truncation [19]. The c.1902 + 5G > A variant had not been reported previously, but in silico analysis predicted it would result in a truncated protein. Blood analysis confirmed an undetectable C2 complement fraction.

Follow-up included regular clinical examinations and monitoring for proteinuria. Hydroxychloroquine was administered to two of the four patients, while antihistamines, used by all patients, were reported to be effective in alleviating itching and angioedema.

SLE-HUVS

P8, P9 and P10 had SLE-HUV. They were 14, 14 and 6 years old at disease onset and 14, 14 and 7 years old at diagnosis, respectively. The mean delay between onset and diagnosis was 9 ± 4 months. All 3 patients presented with recurrent urticaria, angioedema, systemic symptoms and polyarthralgia. SLE-related manifestations included resurgence of seizure activity in a patient with pre-existing epilepsy and new onset of neuropsychiatric features (abnormal headaches, auditory hallucinations, sleep disorders with circadian rhythm reversal, acute cognitive retardation and memory loss) (P8), lupus nephritis class IV (P9), pericarditis (P9), thin pleural effusion and iliac vein thrombosis (P8) and ischaemic brain damage with antiphospholipid syndrome (P10). Laboratory investigations revealed positive ANA, positive anti-dsDNA, low complement levels and positive anti-C1q antibodies in all 3 patients. ANCA was positive in 2 patients (P8 and P10), with a titre of 1:160 and 1:320, respectively (negative anti-myeloperoxidase and anti-proteinase 3). Therapeutic management was guided by SLE manifestations and included hydroxychloroquine, steroids, and immunosuppressive drugs.

Discussion

HUV is a rare and heterogenous disease in children. While most studies have focused either on SLE-HUV or on DNASE1L3 deficiency associated with SLE and/or HUVS, we included all reported French patients with childhood-onset HUV. This series highlights the phenotypic and genotypic diversity observed within one country.

Although the presence of anti-C1q defines the disease in the 2012 revised International Chapel Hill Consensus Conference nomenclature of vasculitis [2], only 6/10 patients in our cohort were positive for anti-C1q antibodies. Due to the multi-centric nature of our study, not all anti-C1q assays were performed in the same laboratory, which may limit the interpretation of our results. However, this proportion of positive anti-C1q antibodies is comparable to the largest adult series (55%) [4], and to the paediatric series reported by Taş et al. (54%) [14]. Notably, none of our 3 patients with S-HUVS associated with DNASE1L3 deficiency had anti-C1q antibodies. Similarly, among the 40 reported children with DNASE1L3 deficiency, anti-C1q antibodies were present in only 5 children [12]. This finding highlights that hypocomplementaemia is not necessarily associated with anti-C1q antibodies, particularly in the monogenic form associated with DNASE1L3 deficiency.

Since 2011, less than 50 children with DNASE1L3 deficiency have been reported, presenting with either HUVS or SLE [12]. Notably, in our study, all 3 patients with S-HUVS—a more severe phenotype—had DNASE1L3 deficiency. In a cohort of 6 children with HUVS, 3 of the 4 children with S-HUVS also had DNASE1L3 deficiency, whereas the remaining 2 patients with NS-HUVS did not [20]. Similar to our study, of the 28 children with HUVS reported by Taş et al., 43% had DNASE1L3 deficiency [14]. The variant observed in siblings P1 and P2 was the same as that previously described by Özçakar et al. [8], while P3, a patient previously reported by Tusseau et al., had a novel homozygous DNASE1L3 c.433G > A variant [12]. These 3 patients underscore the severity of HUVS associated with DNASE1L3 deficiency, which is primarily driven by renal and pulmonary involvement. Among the 40 children with DNASE1L3 deficiency reported by Tusseau et al., renal involvement was more frequent than pulmonary involvement (70% vs. 15%, respectively) [12]. Finally, P1 illustrated the continuum between UV, HUV and HUVS: she initially presented with UV, hypocomplementaemia appeared 1 year later (HUV), and she later developed renal and pulmonary involvement (HUVS).

By contrast, HUVS may manifest as non-systemic, characterised by less severe features such as arthralgia, arthritis, and/or recurrent abdominal pain. In this study, 4 of 10 patients had NS-HUVS, a proportion similar to that reported in the largest adult series, among whom 32% of patients with HUVS had a non-systemic form [4]. Among the patients with NS-HUVS, 3 were positive for anti-C1q antibodies. Interestingly, none of these 3 patients developed renal involvement after 18 to 87 months of follow-up, whereas the presence of anti-C1q antibodies is typically associated with renal involvement in SLE [21].

HUV may also be associated with SLE. In the largest adult series, 25% of HUV cases were associated with autoimmune diseases, primarily SLE [4]. These cases are particularly challenging. SLE, as defined by the 2012 Systemic Lupus International Collaborating Clinics Criteria [15], shares many clinical and biological criteria with HUVS [6], leading to potential overlap and diagnostic difficulties. Among the screened patients with SLE, UV and hypocomplementaemia, we made a critical effort to include only those meeting specific criteria for HUV. Unfortunately, none of these patients underwent a skin biopsy to confirm vasculitis lesions. Some authors have proposed that HUVS is a subtype of SLE or part of a continuum [22]. Davis et al. reported that 50% of HUVS cases evolve into SLE and that 8% of SLE cases are associated with HUV [23]. Similarly, in the largest paediatric series of HUV, all children also had SLE, with 78% initially meeting the criteria for HUV before fulfilling those for SLE [5]. As a result, all patients with HUV should be closely monitored for the potential development of systemic autoimmune diseases, such as SLE.

HUVS is a rare disease that is challenging to diagnose, as highlighted by the long mean diagnostic delays observed in this study. The benign nature of the symptomatology in NS-HUVS seems to contribute to longer diagnostic delays. Nevertheless, HUVS can lead to death in 7% of cases [14]. Early diagnosis is essential to monitor and prevent potentially severe manifestations of the disease, such as renal and pulmonary involvement, and evolution towards SLE. The therapeutic management of HUV is not standardised. Hydroxychloroquine and colchicine appear to be effective first-line therapies in adult patients, but for relapsing or refractory cases, immunosuppressive treatment must be considered on an individual basis [4]. In this study, P1 and P2 achieved remission with rituximab, while antihistamines alone were effective in children with NS-HUVS. By contrast, antihistamines have been reported as ineffective in adults [4].

We reported the third familial form of HUVS associated with DNASE1L3 deficiency, following the two families described by Özçakar et al. [8]. Additionally, we describe an extremely rare case of HUVS associated with C2 deficiency. Of note, 7/10 patients had genetic testing, and a pathogenic genetic variant was identified in 4 of them, accounting for at least 40% of the cohort. However, in the 3 patients with no genetic testing, the disease onset occurred later (after the age of 10 years), while the median age at diagnosis in children with DNASE1L3 deficiency was 4 years [5, 13]. This study highlights genetic causes of paediatric HUVS: physicians should be aware of this entity and systematic genetic testing should be performed. Indeed, beyond improving our knowledge about underlying pathogenic mechanisms, monogenic HUVS may benefit from targeted therapies in a close future and help physician a better tailored standard of care [24, 25].

DNASE1L3 deficiency was first described in 2011 in familial forms of SLE [9]. In 2013, Özçakar et al. reported a homozygous variant (c.289_290delAC) in two distinct families, comprising two and three children, respectively, with HUVS ± SLE [8]. DNASE1L3 is a secreted endonuclease responsible for the clearance of DNA associated with apoptotic cellular debris [26]. Defective clearance of this DNA, which is associated with microparticles, appears to play an important role in activating DNA-specific autoimmune responses in murine models of DNASE1L3 deficiency [10, 11]. The activation of endosomal TLR7 and TLR9 in B cells—and likely in innate immune cells—is crucial for the generation of SLE features in DNASE1L3-deficient mice [11]. Unlike deficiencies in intracellular DNases such as DNASE2 and TREX1, which cause severe diseases in both humans and mice mediated by aberrant and chronic type I interferon (IFN-I) production (interferonopathies) via intracellular DNA sensing by the cGAS/STING pathway, DNASE1L3 deficiency in humans only triggers transient IFN-I production [12]. Moreover, while abrogation of the IFN-I receptor prevents disease caused by TREX1 [27] and DNASE2 deficiency [28] in mice, it only ameliorates SLE features in DNASE1L3-deficient mice without affecting the occurrence of DNA-specific autoreactivity [11]. DNASE1L3 deficiency therefore causes a condition distinct from conventional type I interferonopathies, and may benefit from different therapeutic management [29].

Furthermore, the pathogenesis of HUVS associated with DNASE1L3 deficiency appears to differ from that of classic HUVS. The first difference is the very low – none in our study and only 5/40 children in the review by Tusseau et al. – presence of anti-C1q [12]. The observation of a glomerular vasculitis without complement deposits in P1 also suggests that organ damage in HUVS associated with DNASE1L3 deficiency may not be exclusively due to immune complex deposition. Nevertheless, the presence and characteristics of deposits observed in kidney biopsies are poorly documented in the literature. In three new cases of DNASE1L3 deficiency, recently reported by Volpi et al., the authors detected glomerular expression of the IFN-I-induced human myxovirus resistance protein 1 suggesting that DNASE1L3-related glomerulonephritis could be included in the spectrum of IFN-I-mediated kidney disorders [30]. Another distinctive feature of DNASE1L3 deficiency is the presence of ANCA. Their presence was extremely rare in the large series of adults with HUV [4], and not detailed among the 28 children with HUV reported recently [14], whereas ANCA were found in almost all children with DNASE1L3 deficiency [12], potentially leading to ANCA-associated vasculitis-type kidney or lung damage [31].

The identification of C2 deficiency in P5 is particularly noteworthy. To our knowledge, only two cases of C2 deficiency associated with HUV were previously reported: one patient with SLE-HUV in 1990 [32], and a 5-year-old boy with NS-HUV in 2016 [33]. Of note, a paediatric HUVS case associated with recurrent infections and complement factor 1 deficiency as also been previously described [34]. P5 had a typical form of NS-HUV that began at the age of 5 years, presenting with leukocytoclastic vasculitis on skin biopsy and positive ANA, ANCA and anti-C1q antibodies. Over 9 years of follow-up, P5 did not develop renal or pulmonary involvement, or SLE. C2 deficiency is often associated with recurrent bacterial infections and systemic autoimmune diseases, primarily SLE; cutaneous vasculitis has also been reported [35–39]. These data support the importance of genetic testing even in the least severe and apparently isolated forms of HUVS.

Conclusion

This series of 10 children with HUV highlights the phenotypic diversity of this rare paediatric disease. Therefore, the classification of HUV as a distinct entity remains questionable. Indeed, HUV appears rather to be a heterogeneous syndrome of potentially variable pathogenic origin, which may overlap: monogenic disorders, immune complex-mediated lesions, and SLE, or even ANCA-associated vasculitis, pathogenesis. Children with S-HUVS associated with DNASE1L3 deficiency did not have anti-C1q antibodies and exhibited a more severe phenotype, characterised by renal and pulmonary involvement. Less severe forms (NS-HUVS) were mostly anti-C1q positive and limited to cutaneous and articular involvement. The extremely rare case of NS-HUVS associated with C2 deficiency further underscores the genotypic diversity of this disease. This study highlights genetic causes of paediatric HUVS: physicians should be aware of this entity and systematic genetic testing should be performed. Early diagnosis of HUV is crucial to mitigate and/or prevent severe renal and pulmonary manifestations, and to monitor for the potential development of SLE.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

CB collected data and wrote the manuscript. MB, GB, and MT were involved in genetic analyses. IM, AB, AL, LE, VG, EJ, DL, OR, and PP were involved in the clinical care of the patients and provided data for the study. VS and SL performed DNASE1L3 activity assay. JG helped with data management, edited, and reviewed the manuscript. IM, VS, MB, GB and PP reviewed the manuscript. All authors read and approved the manuscript. All group members fulfil ICMJE criteria for authorship.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Disclosure statement

All authors have agreed to the disclosure of this document.

Competing interests

The authors declare no competing interests.