Blau syndrome is a monogenic granulomatous autoinflammatory disease presenting with polyarthritis, rash, panuveitis and systemic manifestations resulting from gain-of-function point mutations at or near the NOD domain of the NOD2 gene.
Blau syndrome (BS) is an inflammatory disorder considered part of the spectrum of pediatric sarcoidosis and as such primarily characterized by the presence of non-caseating epithelioid giant cell granulomas in a variety of tissues and organ systems. The clinical phenotype associated with an autosomal dominant inheritance pattern was described in 1985 (Blau 1985). In 2001 a mutation in the nucleotide-binding oligomerization domain 2/caspase activation recruitment domain 15 (NOD2/CARD15) was found in four patients with Blau syndrome and with this discovery the first monogenic autoinflammatory granulomatous disease was described (Miceli-Richard et al. 2001). Blau syndrome (BS) and Early Onset Sarcoidosis (EOS) constitute the familial and sporadic forms of this pediatric disease characterized by a triad of polyarthritis, uveitis and rash. Visceral and vascular manifestations beyond the classic triad have since been documented in 29–48% of patients with BS ratifying its true systemic nature (Rose et al. 2009, 2015).
BS is a rare disease and epidemiologic studies focusing specifically on its prevalence do not exist. An approximation can be gathered by looking at sarcoidosis registries. The Danish National Registry included 48 children within a cohort of 5536 patients with sarcoidosis, resulting in a calculated overall incidence for childhood sarcoidosis of 0.29/100,000/year. The incidence ranged from 0.06/100,000/year for children below 5 years old to 1.02/100.000/year for children 14–15 years old (Byg et al. 2003). It is in the younger age group where most cases with BS are found.
Blau syndrome shows an autosomal dominant inheritance which matches the postulated gain-of-function mutation in NOD2. Over the years, a growing number of genetic mutations of NOD2 have been published and reported in the Infevers Registry (Infevers: an online database for autoinflammatory mutations 2017). Substitutions R334W (Arginine to Glutamine in position 334) and R334Q (Arginine to Tryptophan) are by far the most common. More recently, using targeted deep NOD2 sequencing, germline mosaicism or gonosomal somaticism were found in patients who were negative for NOD2 mutations by Sanger sequencing (De Inocencio et al. 2015). NOD2 mutations can be seen in patients with the complete clinical triad but also in both incomplete and expanded phenotypes. Incomplete penetrance in asymptomatic carrier status has rarely been reported (Saulsbury et al. 2009).
The NOD2 gene encodes a 1,040 amino-acid protein composed of three main functional domains namely two amino-terminal Caspase Recruitment Domains (CARDs), a central Nucleotide Binding Oligomerization Domain (NOD/NACHT), and carboxyterminal Leucine-Rich repeats [LRRs]. The NOD2 protein is a member of the family of NOD-like receptor cytosolic proteins [NLRs] involved in pathways of inflammation, apoptosis and phagocytosis. The two amino-terminal CARD domains of NOD2 have an important role in the mediation of nuclear factor NFкB activation and secretion of pro-inflammatory cytokines, resulting from CARD-CARD interactions between NOD2 and a pivotal downstream kinase protein receptor-interacting protein kinase 2 [RIP2]. The centrally located NOD domain mediates self-oligomerization of NOD2 followed by downstream activation of effector molecules. The LRR region is structurally related to the LRR regions of the Toll-like receptors which are pattern recognition molecules of the innate immune system, sensing and binding molecular motifs specific to pathogens. NOD2 like other NLR proteins occurs in two states: a tense comma shaped auto-inhibited state, and a relaxed NOD domain exposure state after ligand engagement. NOD domain exposure is a pre-requisite for NOD2 oligomerization and downstream pathway activation (Boyle et al.2014). Hydrophobic forces and salt bond interactions within the four subdomains of the NOD domain as well as ADP binding maintain the “tight” inactive state (Maekawa et al. 2016). NOD2 oligomerization renders a CARD domain scaffold allowing for interaction with a CARD-containing RIP2 kinase and downstream pathway activation.
The downstream effects of NOD2 auto-activating mutations associated with BS, and their relationship with both granuloma formation and clinical phenotype are not yet understood. Consistent with the autosomal-dominant inheritance pattern and according to early experimental work, NOD2 mutations associated with BS are gain-of-function variants. Transient transfection assays performed in vitro using plasmids with powerful promoters that overexpress NOD2 have found that mutations associated with Blau syndrome cause excessive NF-κB and MAPK activation compared to the wild-type form of NOD2 (Kanazawa et al. 2005). Yet, experiments using patients’ circulating mononuclear and asymptomatic R314Q-knock-in mice show attenuated cytokine production in response to MDP (Dugan et al. 2015). These apparently contradictory data, one could conceive that the gain-of-function effect is not demonstrable in human PBMCs due to a phenomenon of attenuation and/or unknown modulating factors. Although RIP2 activation following NOD2 oligomerization in BS is well-documented, one should bear in mind that there are more than 30 proteins binding NOD2 with different degrees of affinity. Several of the NOD2 binding partners directly influence autophagy. The role of NOD2 in autophagy is of great interest since many of the interacting proteins bind NOD2 at the cytoskeleton and cell membrane where the cell fusion machinery involved in multinucleation and granuloma formation.
Blau syndrome features a clinical phenotype of polyarthritis, dermatitis and uveitis. In recent years, because of the availability of genetic testing, a more protean clinical picture than initially conceived is being unveiled.
The initial manifestations include the typical exanthema followed within months by a symmetrical polyarthritis. Ocular involvement tends to occur later in the disease course.
The rash varies in color from pale pink with varied degrees of tan to intense erythema. The lesions appear on the trunk mainly dorsally and later extend to the face and limbs with accentuation of the tan color on extensor surfaces, where it may become scaly brownish over time. The lesions are tiny (5–7 mm), round, and barely palpable. At onset, the rash often shows a very fine desquamation, which may lead to confusion with atopic dermatitis. Over the course of the years the rash waxes and wanes. With time, the desquamation predominates, and, in adolescence, it may mimic ichthyosis vulgaris.
Subcutaneous nodules, often located in the lower limbs, are the second most common dermatological manifestation and may be clinically indistinguishable from erythema nodosum. The nodules are mildly tender and resolve without atrophy or pigmentation, even in patients with recurrent episodes. Erysipelas-like lesions have been observed as well, and in one case an urticarial rash showed typical histological features of leukocytoclastic vasculitis (Rose et al. 2006).
The majority of patients present with a polyarticular symmetrical or additive arthritis, affecting large and small peripheral joints and tendon sheaths. The most frequently joints involved comprise wrists, knees, ankles and PIP joints. A characteristic feature of both synovitis and tenosynovitis is the exuberance of the swelling. The distal flexor tendons of the digits, the extensor and peroneal compartments, and the flexor groups of the carpus can reach significant size. The synovial outpouching can acquire a cystic appearance in the dorsum of carpus and tarsus. Despite the prominent “boggy” synovitis, pain and morning stiffness appear to be moderate and are overall well tolerated. Except for the proximal interphalangeal joints, where a characteristic flexion contracture described as “camptodactyly” can be seen, the range of motion is relatively well preserved, at least in childhood. The course of the arthritis is variable, and erosive changes are mostly rare and modest. However, limited joint mobility and joint contractures may develop with time; ulnar deviations, wrist subluxations, and joint space narrowing have been described (Rose et al. 2006).
An insidious granulomatous iridocyclitis and posterior uveitis can evolve into a severe destructive panuveitis. Of the clinical triad components, the ocular disease exhibits the most somber functional prognosis. It tends to start within the first 2 years of disease as an asymptomatic uveitis. Over time characteristic iris nodules, focal synechiae and clumpy keratic precipitates at the limbus appear, and cataract and increased intraocular pressure ensue. Nodules may also occur in the conjunctivae and, in this location, offer an early biopsy site and diagnostic clue.
Visceral and Systemic Involvement
It has become apparent that the clinical phenotype of BS is not restricted to the classic triad. According to an ongoing prospective cohort study, systemic and visceral involvement affects 48% of patients with BS (Rose et al. 2015). A myriad of clinical manifestations including granulomatous and interstitial nephritis, chronic renal insufficiency, small vessel vasculitis, interstitial pneumonitis, peripheral and mediastinal (excluding hilar) lymphadenitis, pericarditis, cranial neuropathy (VII cranial nerve), and parotitis have been documented in recent studies (Rose et al. 2009).
Takayasu-like arteritis was recently reported in a girl with BS (Khubchandani et al. 2012). Arterial hypertension with normal digital vascular imaging was observed in 25% of patients from an international registry (Rose et al. 2009).
Peripheral blood cell counts are usually within normal limits, although mild anemia, leucopenia, or lymphopenia can be seen. Levels of C-Reactive Protein were only mildly elevated compared to healthy volunteers, and did not correlate with articular disease activity in a BS cohort study (Rose et al. 2015). Elevation of angiotensin converting enzyme (ACE) is not consistent, and the value of serum ACE levels in diagnosing and managing BS remains unclear. Hypercalciuria and hypercalcemia result from overproduction of 25-hydroxyvitamin D-1 α-hydroxylase by macrophages in granulomas. Hypercalciuria can lead to nephrocalcinosis and nephrolithiasis, both complications being documented in the international prospective BS cohort (Rose et al. 2015).
The diagnosis of BS rests on genetic confirmation of a NOD2 mutation in the context of either a typical clinical phenotype and/or the demonstration of characteristic non-caseating granulomatous inflammation.
Typical non-caseating epithelioid and multinucleated giant cell granulomas can be documented in biopsies of skin, synovium, lymph node, kidney or liver. The skin biopsy has shown the best yield among patients with the classical granulomatous dermatitis. A synovial membrane biopsy can offer a good alternative particularly in patients whose rash has resolved or appears inactive. Blau granulomas display a distinct morphology characterized by large polycyclic granulomas with dense lymphocytic coronas. They reflect an exuberant inflammatory response, which is in line with a gain-of-function mutation in NOD2. Using immunohistochemistry, a predominance of CD68+ macrophages and CD4+T lymphocytes and an abundant inflammatory cytokine expression in situ is typically observed (Janssen et al. 2012).
NOD2 mutations were found in 98% of the patients of the International Pediatric Granulomatous Arthritis Registry exhibiting the classic triad phenotype with either a sporadic or a familial form (Rose et al. 2006).
Hand X- Rays show a symmetrical non-erosive arthropathy with a number of characteristic dysplasia-like bone changes. Some of the most frequently observed deformities including a biconcave radius, carpal crowding, a short plump distal ulna and a thin second metacarpal diaphysis are characteristic and may allow recognition of BS on a single wrist X-Ray view. The radiographic picture which we dubbed the “Blau hand” is very different from the “rheumatoid hand” (Rose et al. 2015).
The diagnosis of BS in a child with granulomatous inflammation requires a concerted effort to exclude chronic infections, notably mycobacterial and fungal, by appropriate staining and cultures.
In geographic areas endemic for tuberculosis, granulomatous arthritis should raise the suspicion of TB. Monoarticular granulomatous synovitis can be seen in patients with foreign body arthritis. Penetration of thorns from Yucca plants, sea urchin spines and other inert foreign bodies are not that unusual in exposed children and should be suspected in the appropriate clinical scenario.
A challenging differential diagnosis for BS polyarthritis with uveitis is polyarticular juvenile idiopathic arthritis (JIA). In addition, adults with BS may be incorrectly diagnosed as rheumatoid arthritis or spondyloarthropathy.
BS needs to be differentiated from other systemic inflammatory disorders associated with granulomatous inflammation in children, such as Crohn’s disease. Although BS does not affect the gastrointestinal tract some extra-intestinal manifestations could be confusing including arthritis, uveitis, hepatitis, cutaneous vasculitis and erythema nodosum. Patients with granulomatosis with polyangiitis (Wegener’s) often have granulomatous inflammation of the upper respiratory tract, a biopsy will reveal signs of small-vessel vasculitis.
Large vessel vasculitis can be a presentation of Blau syndrome therefore among patients with abdominal aortitis, renal artery stenosis, Takayasu’s like syndrome and aortic root disease, Blau syndrome needs to be considered (Khubchandani et al. 2012).
Various primary immunodeficiency disorders can present with granulomatous inflammation without an identifiable infectious cause, and should be excluded by evaluation of neutrophil function, analysis of circulating lymphocyte subsets and serum levels of immunoglobulins (Rose et al. 2014).
Evidence-based data on the optimal treatment for Blau are nonexistent. Moderate- to low-dose daily corticosteroid therapy is effective to control uveitis and joint disease, but the side effects of prolonged use may become unacceptable. In a prospective Blau cohort study, more than two-thirds of Blau patients received medical therapy for several years, often combining systemic steroids, immunosuppressive and/or biologic drugs to control both uveitis and arthritis (Rose et al. 2015). Methotrexate at a dosage of 10–15 mg/m2 once weekly reportedly was effective in suppressing articular disease activity and may be steroid sparing. TNF antagonists were the most commonly used biological therapy. Infliximab and Adalimumab were found to control chronic arthritis and visceral manifestations in a number of patients; however, the effect on uveitis activity is less convincing (Rose et al. 2015). A good response to IL-1 inhibition with Anakinra was reported in a single case (Arostegui et al. 2007) and clinical benefit on refractory uveitis in a 4-year old boy with Canakinumab for 6 months has also been reported (Simonini et al. 2013). Tocilizumab has been used in isolated cases, yet at present its efficacy remains unknown.
Antihypertensive medication may be required in patients who developed arterial hypertension with or without obvious renal involvement. ACE inhibitors have been effective in the few documented cases (Rose et al. 2015).
In the prospective BS cohort study, articular and ocular disease were still active after more than 10 years of systemic therapy. At the baseline evaluation, active ocular inflammation was seen in more than one third of patients and was associated have with moderate to severe visual impairment in 27% and 15% of patients respectively. There was no decrease in inflammatory activity and a progressive loss of visual acuity during 3 years of follow-up (Rose et al. 2015). Arthritis seems nondestructive, especially during the first years, but as the disease progresses, flexion deformities, camptodactyly, and less frequently erosions can be observed. Persistent joint swelling is common, with active arthritis in 70% of patients and a median joint count of 15 in patients with more than 10 years of evolution. Twenty eight percent of patients graded their functional disability as moderate or severe (Rose et al. 2015). Severe hypertension in four patients and visceral involvement, including glomerulonephritis with renal failure (one patient) and interstitial pneumonitis (one patient) was seen among 45 participants in our international retrospective registry (Rose et al. 2009). These findings underline the need for careful surveillance throughout the disease course. Anti-TNF therapy was effective in controlling one case of Blau glomerulonephritis (Rose et al. 2009) and increase in the dose of corticosteroids was effective in controlling interstitial pneumonitis in another patient (Rose et al. 2009). Conversely, pulmonary arterial hypertension was the cause of death in one patient at the age of 23 years (CDR, personal observation).
- Arostegui JI, Arnal C, Merino R, Modesto C, Antonia CM, Moreno P, et al. NOD2 gene-associated pediatric granulomatous arthritis: clinical diversity, novel and recurrent mutations, and evidence of clinical improvement with interleukin-1 blockade in a Spanish cohort. Arthritis Rheum. 2007;56:3805–13.CrossRefPubMedGoogle Scholar
- Boyle JP, Parkhouse R, Monie TP. Insights into the molecular basis of the NOD2 signaling pathway. Open Biol. 2014. https://doi.org/10.1098/rsob.140178.
- Infevers: an online database for autoinflammatory mutations. http://fmf.igh.cnrs.fr/ISSAID/infevers/lastupdated 10/29/2014. Accessed 8 July 2017.
- Janssen CE, Rose CD, DeHertogh G, Martin T, Badeer-Meunier B, Cimaz R, Harjacek M, Quartier P, Cate RT, Tomee C, Desmet VJ, Fischer A, Roskams T, Morphological WCH. Immunohistochemical characteristics of granulomas in the NOD-2-related Pediatric granulomatous disorders Blau syndrome and Crohn’s disease. J Allergy Clin Immunol. 2012;129:1076–84.CrossRefPubMedGoogle Scholar
- Maekawa S, Ohto U, Shibata T, Miyake KShimizu T. Crystal structure of NOD2 and its implications in human disease. Nature Comm. 2016. https://doi.org/10.1038/ncomms11813.
- Rose CD, Pans S, Casteels I, Anton J, Bader-Meunier B, Brissaud P, et al. Blau syndrome: baseline data from a prospective multicenter cohort study of clinical, radiological and functional outcomes. Rheum (Oxford) 2015; 54:1008–16.Google Scholar