Polyarteritis nodosa and deficiency of adenosine deaminase 2– shared genealogy, generations apart

Abstract

Polyarteritis nodosa (PAN) is a systemic necrotizing vasculitis that predominantly affects medium-sized arteries. With the establishment and refinement of vasculitis nomenclature and diagnostic criteria, clinical findings of PAN and distinguishing features from other vasculitides are now well characterized. Although PAN typically manifests in adulthood, cohort studies in pediatric patients have shaped our understanding of childhood-onset PAN. The paradigm of childhood-onset PAN changed considerably with the landmark discovery of deficiency of ADA2 (DADA2), a monogenic cause of vasculitis that is often indistinguishable from PAN. Testing for DADA2 has provided an explanation to numerous challenging cases of familial PAN and early-onset PAN around the world. The ability to distinguish DADA2 from classic PAN have important therapeutic implications as tumor necrosis factor inhibitors have demonstrated remarkable efficacy in the treatment of DADA2. In this review, we will discuss our current understanding of PAN and DADA2 and highlight similarities and differences between these vasculitides.

Introduction

Systemic vasculitis are a group of heterogeneous disorders that share the pathologic finding of blood vessel wall inflammation. Clinical manifestations of systemic vasculitis are largely due to chronic inflammation, vascular damage resulting in hemorrhage and/or vascular occlusion leading to ischemia and infarction of downstream tissues and organs [1]. Polyarteritis nodosa (PAN) is a systemic necrotizing vasculitis of medium-sized arteries with occasional small artery involvement [2]. Like other systemic vasculitides, clinical manifestations of PAN encompass a broad spectrum ranging from non-specific constitutional symptoms to organ-specific involvement affecting almost every organ system.

PAN is typically a sporadic disease but can present as a secondary process to infection (e.g. hepatitis B), drug reaction or another inflammatory condition such as familial Mediterranean fever (FMF) [3]. Few susceptibility alleles have been linked to PAN and familial cases are uncommon [4, 5]. While PAN is considered a disease of adulthood with a median age of disease onset around the 4^th to 5^th decade of life, childhood-onset PAN is increasingly recognized with some variations in clinical manifestations compared to the adult disease [6, 7].

Our view of childhood-onset PAN changed considerably with the discovery of deficiency of adenosine deaminase 2 (DADA2) [8, 9]. Many cases of PAN in children, particularly those with early disease onset or affected family members, were found to have biallelic mutations in ADA2. Although much remains to be discovered on how absence of ADA2 translates to the development of vasculitis, DADA2 is now a well-recognized mimic of PAN [10]. However, the disease manifestations of DADA2 extend beyond vascular inflammation and the treatment approach is not the same as PAN [11].

In this article, we will review the current understanding of classic PAN and DADA2. We will compare diagnostic considerations, clinical features, and treatment approaches of both entities and highlight the distinguishing features.

Historical perspective and current understanding of PAN and DADA2

We will briefly outline major milestones that set the foundation for our understanding of PAN and DADA2. Readers are encouraged to read more detailed reviews on the history of PAN and vasculitis nomenclature / classification [12–18].

The initial description of pathologic features of PAN can be traced back to the mid 19^th century. In 1852, Rokitansky provided the first pathologic description of numerous arteries with abundant aneurysmal nodules from the autopsy of a young man who died shortly after he presented with fever, abdominal pain and bloody diarrhea [19]. Kussmaul and Maier in 1866 described a 27-year-old man with constitutional symptoms, myalgia, mononeuritis multiplex and abdominal pain [20]. The patient experienced rapid clinical decline and autopsy was notable for “nodular thickening of countless arteries”. The authors coined the term “periarteritis nodosa” to describe the perivascular inflammation in the adventitia and media layers, with sparing of the vessel intima [20]. In the early 1900s, Ferrari and Dickson introduced the term polyarteritis acuta nodosa to describe their observation of transmural vascular inflammation with nodulosis affecting multiple vessels [21, 22].

With better understanding of other systemic vasculitides, the definition of PAN was narrowed to arteritis affecting medium and small arteries without involvement of arterioles, capillaries, or venules. The standardized nomenclature system established by the international Chapel Hill Consensus Conference (CHCC) in 1994 created clear distinction between PAN and microscopic polyangiitis (MPA), a related necrotizing vasculitis that primarily affects small vessels [12]. The evolution of vasculitis nomenclature and classification schemes have also distinguished “classic” PAN from organ-specific vasculitis and secondary PAN associated with infection, drugs and other systemic diseases.

The prevalence of various systemic vasculitides differs greatly between adults and children [23]. IgA vasculitis / Henoch-Schönlein purpura (HSP) and Kawasaki disease (KD) comprise the majority pediatric vasculitis cases while PAN in children is generally rare. Some early cases of “infantile PAN” in the literature with coronary artery aneurysms in young children were difficult to distinguish from KD, a medium-sized vessel vasculitis with onset largely restricted to childhood [24]. Although classic PAN in children shares many features of the adult disease, the early diagnostic criteria for PAN were not validated in children [23]. To address the need to better understand vasculitis in children, the 2008 Ankara Consensus Conference formalized the diagnostic criteria for childhood-onset vasculitides including PAN [17, 18].

The etiology of classic PAN remains unclear. Unlike vasculitis associated with antineutrophil cytoplasmic antibodies (ANCA), PAN does not possess strong association with human leukocyte antigen (HLA) polymorphisms or pathognomonic autoantibodies that are typical for an autoimmune process. There is also a paucity of literature on the role of B and T lymphocytes in PAN [25]. Neutrophil infiltration and fibrinoid necrosis in the vessel wall support activation of the innate immune system, potentially suggesting that PAN could represent part of the autoinflammatory spectrum. This view is supported by the discovery of DADA2, a monogenic autoinflammatory syndrome that recapitulates many hallmark features of PAN.

In 2014, DADA2 was described by two groups via whole exome sequencing (WES) of patients with early-onset vasculitis and/or familial PAN [8, 9]. These initial studies together identified biallelic mutations in the gene CECR1 (cat eye syndrome critical region candidate 1) in more than 30 patients from around the world, many of whom were previously diagnosed with PAN. This gene is not responsible for cat eye syndrome and was later found to encode the protein ADA2 [26]. The gene nomenclature was changed to ADA2 in 2018.

Subsequent studies from other centers showed that up to ¼ of childhood PAN cases were explained by DADA2 [27, 28]. While DADA2 usually presents during childhood, adult-onset cases have been described [29]. Beyond PAN-like vasculitis, the clinical spectrum of DADA2 also includes pure red cell aplasia, bone marrow failure syndrome, humoral immunodeficiency and lymphoproliferative disease [11]. Environmental factors, genetic background and epigenetic modifications likely play important contributory roles, because individuals with identical mutations may diverge with respect to clinical manifestations and disease severity [30, 31]. Some patients experience multiple strokes in the first few years of life while others remain completely asymptomatic through adulthood, without even serological evidence of inflammation [32]. To date, more than 200 cases of confirmed DADA2 have been described in the literature.

The function of ADA2 and the pathophysiology of DADA2 are topics of active investigation. In the initial description of DADA2, Zhou et al. proposed that aberrant polarization of monocytes and macrophages to the proinflammatory M1 phenotype in the absence of ADA2 resulted in cytokine production and endothelial damage [8]. Recent work by Carmona-Rivera and colleagues showed that enhanced neutrophil extracellular trap formation secondary to the accumulation of adenosine drives the production of tumor necrosis factor (TNF) by M1 macrophages [33]. Interestingly, upregulation of type I interferon signaling in DADA2 patients was observed by several investigators [34–36], raising the possibility that DADA2 may be in the spectrum of the “interferonopathy” family of autoinflammatory diseases. Further research is needed to determine how lack of ADA2 translates into systemic vasculitis and other disease features in DADA2.

Diagnostic evaluation for PAN and DADA2

The ACR diagnostic criteria for PAN established in 1990 captured both clinical and laboratory features of the disease (Table 1) [37]. These criteria did not distinguish MPA from PAN and did not account for ANCA testing. These important limitations were addressed in the definition of PAN by the CHCC [12]. Incorporating both ACR criteria and CHCC definitions along with histologic and radiographic findings, Watts and colleagues established a diagnostic algorithm to delineate PAN, MPA and granulomatosis with polyangiitis (GPA) [38]. The French Vasculitis Study Group also put forth a set of positive and negative discriminant criteria to aid the diagnosis of PAN [39].

Table 1.

Diagnostic criteria for PAN in adults and children

1990 ACR criteria for the classification of polyarteritis nodosa [37]*

1) Weight loss of 4 kg or more

2) Livedo reticularis

3) Testicular pain/tenderness

4) Myalgia or leg weakness/tenderness

5) Mononeuropathy or polyneuropathy

6) Diastolic blood pressure greater than 90 mm/Hg

7) Elevated blood urea nitrogen or creatinine level unrelated to dehydration or obstruction

8) Presence of hepatitis B surface antigen or antibody in serum

9) Arteriogram demonstrating aneurysms or occlusions of the visceral arteries

10) Presence of polymorphonuclear neutrophils in a biopsy specimen from a small- or medium-sized artery

2008 EULAR/PRINTO/PRES childhood PAN criteria [17, 18]

A systemic illness characterized by the presence of either a biopsy showing small and medium-size artery necrotizing vasculitis

OR

Angiographic abnormalities demonstrating aneurysms or occlusions

plus at least two of the following:

1) Skin involvement (livedo reticularis, subcutaneous nodules, ulcerations and necrosis)

2) Myalgia or muscle tenderness

3) Hypertension (> 95% for height)

4) Peripheral neuropathy

5) Renal involvement

^*presence of 3 or more criteria is considered diagnostic for PAN

The collaborative efforts by EULAR (European League Against Rheumatism), PRINTO (Paediatric Rheumatology International Trials Organisation) and PRES (Paediatric Rheumatology European Society) established validated diagnostic criteria for childhood PAN (Table 1) [16, 18]. A major distinguishing feature from the ACR criteria for adult PAN is the requirement of biopsy or angiographic studies to confirm the diagnosis. Biopsy of clinically affected tissues in PAN typically shows necrotizing vasculitis of medium-size arteries. Angiographic studies including computed tomography, magnetic resonance imaging or conventional arteriography can be utilized to demonstrate the presence of aneurysms or occlusions in the affected vessels [2]. Other notable differences of childhood PAN criteria include 1) expansion of cutaneous PAN findings to incorporate subcutaneous nodules, skin ulceration, infarction and necrosis; and 2) removal of criteria for weight loss and for hepatitis B seropositivity, reflecting the difficulty of applying an absolute weight-loss cutoff in children of different ages as well as the rarity of concurrent hepatitis infection in childhood PAN.

As a monogenic syndrome, DADA2 can be diagnosed in one of two ways: genetic testing to identify ADA2 mutations or determination of ADA2 enzyme activity in the peripheral blood [10, 40]. Many commercial next generation sequencing panels for evaluation of primary immunodeficiencies or periodic fever syndromes include ADA2. WES is another common approach for DADA2 diagnosis, especially in cases with atypical presentation. In some cases, the mutations (often large insertions, deletions or duplications) can be missed by sequencing, requiring specialized assays designed to detect these cryptic mutations [41, 42].

Because ADA2 is secreted into the circulation by activated monocyte and macrophages, ADA2 enzymatic activity can be readily measured in the peripheral blood and in the supernatant of cultured monocytes / macrophages. Available spectrophotometric assays typically detect the release of ammonia from the deaminase reaction of converting adenosine to inosine (Figure 1A) [8, 43]. Alternatively, high performance liquid chromatography can be used to quantify the production of inosine and hypoxanthine downstream of adenosine deamination [27]. In both methods, the addition of an ADA1 inhibitor is required for specific measurement of ADA2 activity. Figure 1B illustrates results of ADA2 enzyme activity measurement using a spectrophotometric assay that quantifies the adenosine-dependent generation of ammonia. Levels of ADA2 activity are generally higher in healthy children compared to adults but vary considerably within each group (Figure 1B) [44]. Some healthy individuals have low ADA2 activity levels that overlap with the range for carriers with one ADA2 mutation. Very low to absent ADA2 activity is diagnostic for DADA2. The authors recommend using ADA2 activity levels as an initial screening test for evaluation of DADA2 whenever possible because enzymatic tests are considerably faster and less costly compared to genetic testing. In addition, some variants may be of uncertain significance, such that direct measurement of ADA2 enzymatic activity provides a simultaneous gauge of protein abundance and function.

Figure 1.

Methods of detecting ADA2 activity. A) Schematic of the reaction catalyzed by ADA2. Enzymatic assays for determination of ADA2 activity typically capture the release of ammonia from the deaminase reaction while high performance liquid chromatography (HPLC) can be used to quantified inosine and the downstream product hypoxanthine (generated by purine nucleoside phosphorylase). Inhibition of ADA1 by the specific inhibitor EHNA is required to accurately quantify ADA2 activity. B) Representative results of ADA2 activity in healthy controls, carriers and DADA2 patients measured by an enzymatic assay coupled to the release of ammonia. The number of subjects, median ADA2 activity and interquartile range for each group are: healthy children (ages 1 – 18, n = 100; 12.6 U/L; 10.4 – 15.5 U/L), healthy adults (ages 19 – 72, n = 100; 10.2 U/L; 7.3 – 13.4 U/L), carriers (n = 10; 4.25 U/L; 3.9 – 4.8 U/L), and DADA2 patients (n = 10; 0.3 U/L; 0.2 – 0.7 U/L).

Genetics of PAN and DADA2

The majority of PAN cases are sporadic and the role of genetics is largely unclear. Genetic association with human leukocyte antigen (HLA)-DR polymorphisms and MEFV gene mutations responsible for FMF have been reported in PAN. In a Brazilian cohort of 29 patients, HLA-DR alleles were linked to disease severity and organ-specific manifestations [4]. However, an earlier study in the United States found an association of HLA-DR in 15 patients with GPA, but not PAN [5]. The discordance between these studies may be due to the limited sample size, but population-specific HLA associations cannot be ruled out.

An increase incidence of PAN was noted in patients with FMF [45]. The MEFV gene was later identified as a susceptibility allele for PAN in Turkey as FMF-related mutations were found in 11 of 29 pediatric PAN patients [46]. Whether this observation is generalizable in other parts of the world where FMF is less prevalent remains to be seen. The incidence of PAN associated with FMF in Turkey also seems to be declining in recent years, possibly due to advances in the treatment for FMF [46].

DADA2 is caused by biallelic mutations in ADA2. To date, modifier genes that influence ADA2 production or activity have not been identified. Since the initial discovery of DADA2, more than 80 disease-associated ADA2 mutations have been identified from over 200 patients described in the literature (Figure 2A). Mutations have been found in every exon of the gene. Highlighting the hereditary nature of DADA2, more than half of patients have homozygous mutations (Figure 2B). Missense variants represent about 80% of mutations while insertions / deletions, nonsense mutations and splice site variants are less common (Figure 2C).

Figure 2.

The genetic landscape of DADA2 summarized from the first 218 cases in the literature. A) Listing of published ADA2 mutations by exon location. Gray areas indicate untranslated regions. B) Comparison of mutation zygosity in DADA2. C) Pie chart illustration of various mutation types in DADA2. D) Listing of the most common mutations in DADA2 based on the number of identified alleles. Cases with homozygous mutations were counted as two alleles.

The most common pathogenic mutations in DADA2 are p.Gly47Arg (G47R) and p.Arg169Gln (R169Q; Figure 2D). G47R is found at high frequency in DADA2 patients from Turkey, Israel and South Asia. Patients with homozygous G47R mutations typically exhibit the classic PAN-like phenotype [9, 43]. In contrast, the R169Q variant likely originated in Northern Europe, and homozygous patients display variable features of vasculitis, severe hematologic defects, and/or immunodeficiency [31]. Although DADA2 is considered a rare disease, it is worth noting that the frequency of R169Q in the general population is about 1 in 2100 individuals (minor allelic frequency 4.74 × 10⁻⁴ in the gnomAD database).

With the expanding clinical spectrum of DADA2, genotype-phenotype correlation is a topic of active investigation. One recent study suggested that mutations in the dimerization domain are associated with vasculitis while those in the catalytic domain are linked to the pure red cell aplasia phenotype [43]. Our studies of variants reported worldwide have suggested instead that missense mutations with greater residual enzymatic activity are associated with vasculitis, whereas missense variants with minimal residual function and mutations that cause disruption of protein translation (e.g. nonsense and frameshift) are commonly found in patients with severe hematologic manifestations [76]. Further work is needed to define how these genotype-phenotype correlations reflect the pathophysiology of DADA2.

Clinical and laboratory features of PAN and DADA2

As a vasculitis affecting medium- and small-sized arteries, PAN can involve almost every organ system. Table 2 compares the approximate frequency of clinical features in adult PAN, childhood PAN and DADA2 summarized from large case series and reviews [6, 7, 11, 47–49].

Table 2.

Comparison of clinical features in PAN and DADA2 ^*

	Adult PAN	Childhood PAN	DADA2
Average age at diagnosis (yr)	50	9	5
Female (%)	35	50	50
Clinical features
Recurrent fever (%)	60	80	70
Myalgia / arthralgia (%)	60	70	50
Livedo (%)	20	50	70
Cutaneous vasculitis (%)	60	90	80
Neurologic manifestations (%)	60	30	60
CNS involvement (%)	10	20	60
PNS involvement (%)	70	20	20
GI manifestations (%)	30	40	40
Renal manifestations (%)	40	40	20
Hypertension (%)	40	30	20
Laboratory features **
Hemoglobin	↓	-	↓ (50%)
WBC	↑	-	↓ (50%)
Platelets	↑	-	↓ (20%)
IgA	-	-	↓ (40%)
IgM	-	-	↓ (50%)
IgG	-	-	↓ (30%)

Bold print highlights areas of major difference between DADA2 and PAN

^*Data are approximated based on references [6, 7, 11, 47–49]

^**Arrows indicate increased (↑) or decreased (↓) levels compared to healthy controls. The approximate percentage of DADA2 patients with each laboratory feature is provided.

Most patients with PAN present with constitutional symptoms including fever, malaise, fatigue and weight loss [47]. Neurologic manifestations, primarily affecting the peripheral nervous system (PNS; i.e. mononeuritis multiplex, polyneuritis and peripheral neuropathy), are present in the majority of adult patients. Disease of the central nervous system (CNS) such as ischemic stroke, intracranial hemorrhage and cranial nerve deficits is less common. Cutaneous manifestations including livedo, subcutaneous nodule, purpura, ulceration, and necrosis are found in more than 50% of cases. About half of patients experience myalgia and/or arthralgia but arthritis and myositis are less common [7].

The gastrointestinal tract and kidneys are common sites of parenchymal involvement. Gastrointestinal manifestations range from non-specific abdominal pain, nausea and vomiting to visceral artery aneurysm, intestinal infarction and bowel perforation. Renal features of PAN include hematuria, proteinuria, renal artery aneurysm and renal infarct [50]. Hypertension and impaired renal function are also seen but in contrast to vasculitides with small vessel involvement, glomerulonephritis is not a feature of PAN [51]. The lungs are usually spared in PAN, a pattern distinct from GPA and MPA, both small vessel diseases with a preponderance of pulmonary manifestations. Involvement of the cardiovascular, pulmonary and genitourinary system occur in <10% of cases in PAN [50].

The clinical features and their frequency in adult and childhood PAN largely overlap. Notable distinctions between the two groups are more frequent cutaneous involvement and less neurologic complications in children compared to adults [6, 7, 49]. No sex bias is seen in childhood PAN while a male predominance is observed in adult patients. The prognosis of childhood PAN is generally considered favorable compared to the adult disease, with less mortality and morbidity [7].

From clinical manifestations to microscopic examination of biopsy specimens, most features of vasculitis are indistinguishable between PAN and DADA2. In fact, case series of childhood PAN often included patients that were later diagnosed with DADA2 [49]. Demographic features that may help segregate these entities include the age of onset and a family history of PAN. Disease onset tends to be earlier for DADA2, but late presentation of DADA2 and enzyme deficiency in asymptomatic adults have been described [29, 32]. Cutaneous involvement is generally more common in childhood PAN and DADA2 compared to adult PAN. The frequency of livedo ranges from 20% in adult PAN to over 70% of DADA2 patients.

While PNS involvement is predominantly found in adult PAN, the frequency of CNS manifestations (e.g. ischemic stroke and brain hemorrhage) is several fold higher in DADA2 compared to PAN. Many DADA2 patients experience recurrent strokes until they receive the appropriate treatment [52]. Intriguingly, knockdown of ADA2 expression in zebrafish by morpholinos also results in intracranial hemorrhage [8]. The greater risk for ischemic infarct and brain hemorrhage in DADA2 compared to PAN has not been explained. Local production of ADA2 in the CNS, possibly by microglial cells, may have a role in maintaining vascular integrity in the brain.

Laboratory features of PAN including mild anemia, leukocytosis and thrombocytosis generally reflect a state of systemic inflammation [49, 53]. In contrast, DADA2 is associated with a wide spectrum of bone marrow defects (e.g. anemia, neutropenia, lymphopenia, and/or thrombocytopenia) and humoral immunodeficiency (e.g. low IgA, IgG, and/or IgM levels, reduced number of switched memory B cells) [11]. The variable features of bone marrow failure in DADA2 may be related to the proposed role of ADA2 as a growth factor that promotes immune cell proliferation and differentiation [54, 55].

While more data are needed to compare laboratory features of adult and childhood PAN, cytopenias and low immunoglobulin levels favor the diagnosis of DADA2. Consistent with this view, a recent study proposed lower platelet count as a distinguishing feature of DADA2 compared with PAN [43].

Treatment of PAN and DADA2

Aggressive treatment for PAN is necessary because the 5-year survival rate is less than 15% for untreated patients [56, 57]. However, heterogeneity of its clinical manifestations complicates any attempt to define a unified treatment approach. Readers are encouraged to review the latest update on therapeutic options for PAN and treatment algorithms based on disease severity [2, 58]. Since their initial use in 1950, corticosteroids remain the first-line treatment for PAN [59]. However, corticosteroids achieve sustained remission in only approximately 50% of patients [60]. Cytotoxic immunosuppressive agents are therefore widely used for the treatment of PAN, particularly for corticosteroid-resistant cases or those with frequent relapses [61]. Beneficial effects have been reported for cyclophosphamide, azathioprine, mycophenolate mofetil, methotrexate and cyclosporine [62–66]. The prognosis of PAN has substantially improved with the combined use of corticosteroid and cytotoxic immunosuppressive agents, and current 5-year survival exceeds 90% [67].

Biologics have been used in cases refractory to standard therapy, but data remain limited. In case series of adult and childhood PAN, infliximab and other TNF inhibitors (TNFi) were generally effective and well-tolerated [68, 69]. The use of tocilizumab (a monoclonal antibody to IL-6 receptor) in several patients with recalcitrant disease also showed sustained benefit [70]. High-dose intravenous immunoglobulin (IVIG) can be considered for the treatment severe cutaneous polyarteritis nodosa [71, 72]. For patients with hepatitis B- or hepatitis C-associated PAN, anti-viral agents should be regarded as the first-line treatment, but short-term usage of corticosteroids and plasma exchange can be considered [73, 74].

Because of the overall resemblance of DADA2 to PAN, most immunosuppressive agents used for PAN have been trialed in DADA2. There are no head-to-head trials to compare these agents but TNFi have emerged as the standard of care for patients with vasculitis and stroke. After reports of treatment efficacy from early case series [9, 27], Ombrello and colleagues recently demonstrated the remarkable impact of TNFi on preventing strokes in DADA2 patients. In a cohort of 15 DADA2 patients who had experienced 37 combined stroke events prior to TNFi therapy, no additional stroke events were recorded in 733 patient-months of follow up after treatment [52]. The choice of TNF inhibitor does not seem to influence the outcome. Other biologics that target IL-1 and IL-6 have been used in a few patients with variable results, and many patients were subsequently switched to TNFi [8, 9, 41]. Whether all DADA2 patients should be given TNFi is a subject of debate in the field. Our own experience with TNFi is consistent with that of Ombrello, and given the potentially devastating consequences of cerebrovascular compromise, we have a low threshold to start these agents in patients with a confirmed diagnosis of DADA2. Whether asymptomatic family members bearing identical genotypes should be treated as well remains a difficult question, but in appropriate circumstances prophylactic treatment could be considered.

The efficacy of TNFi for non-vasculitis features of DADA2 is less clear. In patients with humoral immunodeficiency, additional immunosuppression may further increase the susceptibility to serious infection, in particular since the response of immune and/or hematologic manifestations to TNFi is typically less robust than of the inflammatory / vascular manifestations. For patients with severe hematologic manifestations, either pure red cell aplasia or bone marrow failure, early hematopoietic stem cell transplant (HSCT) may be the preferred approach if a suitable donor is available. Improvement of almost all aspects of the disease was observed in a cohort of 14 patients who underwent HSCT [75].

Gene therapy and gene editing are being investigated as potential treatment options in DADA2. Enzyme replacement is effective for the treatment of ADA1 deficiency but has so far not been attempted for DADA2. Fresh frozen plasma and some IVIG preparations contain ADA2 but the short plasma half-life of ADA2 limits the utility these options [26, 52].

Conclusion

The history of PAN started with the initial autopsy observation and case description in the mid-1800s. We now have a broad understanding of the clinical features and hallmark pathologic findings of PAN. Our ability to distinguish PAN from other systemic vasculitides has improved with continued refinement of classification schemes and diagnostic criteria. However, our comprehension of the underlying etiology remains incomplete. The discovery of DADA2 has taught us how defects in a single gene can bring out the complex picture of PAN. Ongoing mechanistic studies on DADA2 may help improve our understanding of PAN and define more specific treatment targets.

Highlights:

• PAN and DADA2 are systemic vasculitis that primarily affect medium-sized arteries.

• PAN is diagnosed by established criteria in adults and children while DADA2 is determined by genetic studies and/or enzymatic testing

• DADA2 should be ruled out for familial and early-onset cases of PAN

• PAN and DADA2 are often clinically indistinguishable but presence of hematologic and immunologic defects favors the diagnosis of DADA2

Abbreviations:

ACR

American College of Rheumatology

ADA2

adenosine deaminase 2

CNS

central nervous system

DADA2

deficiency of ADA2

EULAR

European League Against Rheumatism

FMF

familial Mediterranean fever

gnomeAD

Genome Aggregate Database

GPA

granulomatosis with polyangiitis

HLA

human leukocyte antigen

HSCT

hematopoietic stem cell transplant

HSP

Henoch-Schönlein purpura

IVIG

intravenous immunoglobulins

KD

Kawasaki disease

MPA

microscopic polyangiitis

PAN

polyarteritis nodosa

PNS

peripheral nervous system

PRES

Paediatric Rheumatology European Society

PRCA

Pure red cell aplasia

PRINTO

Paediatric Rheumatology International Trials Organisation

TNF

Tumor necrosis factor

TNFi

TNF inhibitor

WES

whole exome sequencing