Monogenic polyarteritis: the lesson of ADA2 deficiency

The enzyme Adenosine Deaminase (ADA) plays a key role in the purine metabolism converting adenosine to inosine and 2′-deoxyadenosine to 2′-deoxyinosine [15].

The two major ADA isoforms are ADA1, whose deficiency is cause of a severe combined immunodeficiency (SCID) [16], and ADA2.

Even if the two proteins have partial structural homology, the two isoenzymes differ in many aspects: the affinity of ADA2 for molecules of adenosine and deoxy-adenosine is about 100 times lower than that of ADA1; consequently, at physiological concentrations of substrate, the deaminase activity of ADA2 is almost absent [17].

While ADA1 is monomeric and acts primarily intracellularly, ADA2 is dimeric and secreted in the extracellular environment where it exerts its main functions. For this reason ADA2 is clearly detectable in the plasma. Finally, while ADA1 is ubiquitally expressed in all cell types, ADA2 is mostly expressed by monocytes and other cells of the myeloid lineage [17].

ADA2 is more stable at high temperatures and the optimum pH for its activity is generally acid (about 6.5), which suggests a specialized role of this enzyme in conditions of hypoxia, inflammation and oncogenesis; in these conditions its deaminase activity is higher [17] (Fig. 2).

The capacity of binding receptors involved in the signal transduction of different pathways (such as proteoglycans), confers to ADA2 a growth-factor like action; for this reason ADA2 is considered an Adenosine Deaminase-related Growth-Factor (ADGF) [17‐19] (Fig. 2).

ADA2 displays also an autocrine activity: the protein, released by activated monocytes, is able to induce monocyte proliferation and macrophage differentiation [20]; CECR1 silencing in myeloid cells is in fact associated to a reduced differentiation of monocytes to macrophages [1]. This activity has been demonstrated to be mediated by the direct binding of cellular receptors, and therefore to be independent from the enzymatic activity [20] (Fig. 2).

ADA2 seems to be also involved in the balance between pro-inflammatory (M1) and anti-inflammatory (M2) monocytes; its absence has been in fact associated with a defect in differentiation of M2 macrophages, which leads to a prevalence of pro-inflammatory M1 cells [1].

Microarray analysis in two DADA2 patients showed a marked up-regulation of neutrophils-expressed genes. Intracellular staining revealed an increased expression of myeloperoxidase (MPO) in peripheral blood mononuclear cells compared to controls [8] (Fig. 2). According with these findings, the assessment of cytokine levels performed in the serum of few described patients revealed an increase of pro-inflammatory cytokines: in the two patients carrying a homozygous deletion of 22q11.1 chromosome increased levels of both IL-1β and TNFα were detected [14], while in another case the detection of IL-6 revealed increased levels [7]. These data are in contrast with the results obtained in the NIH study, in which the cytokine assay performed in the supernatants of the whole blood cell cultured with different stimuli did not reveal any significant difference between patients and healthy donors [1]. Further studies on larger series of patients are therefore needed in order to investigate the cytokines’ pattern in DADA2; in particular the cytokines’ production should be assessed in stimulated PBMCs and should take into consideration the disease activity.

It has been also postulated that the deregulation of purinergic stimulation, due to the decrease of the enzymatic activity of ADA2, may play a pro-inflammatory role. Adenosine is in fact an important signaling molecule that can modulate the inflammatory response; its concentration in tissues is normally low and increases in condition of cellular stress, ischemia or inflammation [21]. The accumulation of adenosine can influence the inflammatory response by binding several receptors that lead to inflammation, tissue damage and fibrosis [21]. However, the plasmatic levels of adenosine and deoxiadenosine in few DADA2 patients has been detected within the normal range [1, 2].

Since hypogammaglobulinemia has been described in some patients, adaptive immunity has been investigated in ADA2 patients. A reduction in the number of memory B cells, terminally differentiated B cells and plasmacells has been described [1, 7]; moreover co-culture experiments have enlightened an increased mortality of B cells [1]. Not univocal results have been detected concerning the T cells function. In fact, while in the NIH study ADA2 mutations seem not to affect T lymphocyte number and activation [1], in a more recent study an increase of regulatory T cells and a decrease of CD8+ and CD4 + memory T cells have been detected in one patient with DADA2 [7]. In addition a reduced number of Th1, Th2 and follicular T helper (Tfh) cells has been observed in the same patient.

The reason why endothelium represents the main target of inflammation in DADA2 is still largely unknown. ADA2 acts as a growth-factor for endothelial cells. In fact, even if it has been demonstrated that endothelial cells do not express CECR1 gene, the deficiency of ADA2 is associated to a damage of vascular endothelium and to an over expression of activation markers [1]. The knockdown animal model for CECR1 gene (zebrafish) displays cerebral haemorrhages without morphologic alteration in the vascular structure; these episodes recovered following the transfection of non-mutant human CECR1 messenger RNA [1]. In the same way, monocytes of patients with DADA2 led to destruction of co-cultured human microvascular endothelial cells [1].

Due to the rarity of the disease, all available data on the pathogenic consequences of ADA2 defect in humans come from few patients; further studies are therefore needed in order to better enlighten the activity of ADA2 in the innate and adaptive immune response and its role in the endothelium homeostasis.

DADA2 can be defined as an inflammatory vasculopathy with a wide range of clinical manifestations, possibly associated with an immunodeficiency of variable severity.

The disease is mainly characterized by chronic or recurrent systemic inflammation with fever and elevation of acute phase reactants, usually associated with different possible skin manifestations, ranging from the most frequent livedo reticularis (Fig. 3) to maculopapular rash, nodules, purpura, erythema nodosum, Raynaud’s phenomenon, ulcerative lesions, digital necrosis [1, 2].

The clinical picture is wide, ranging form a mild disease with a late onset skin-limited involvement to a very severe systemic phenotype (even fatal) with an early onset and a multi-organ involvement (Tables 2 and 3).

In most patients, a neurological involvement, affecting both the peripheral and central nervous system, has been described.

The severity of the CNS involvement is rather variable. In some patients a transitorily ischemic attack (TIA) has been described (with negative cerebral CT and/or MRI), while others developed an ischemic or hemorrhagic stroke (in few cases a ventricular haemorrhage has also been detected). Typically, the strokes associated to DADA2 are lacunar with a wide range of clinical manifestations ranging from clinically silent episodes in few cases, to severe ones leading to a permanent disability [1, 2, 10, 12].

The neuropathy ranges from a transient mononeuritis (such as a cranial nerve transient paralysis) to a permanent polyneuropathy; moreover, few patients suffered from optic neuritis. In few cases, persistent neurosensorial hearing loss has also been described [1, 2, 12].

Most patients have gastrointestinal manifestations: abdominal pain, significant weight loss, chronic gastritis, hepatomegaly, splenomegaly, portal hypertension, bowel perforation or stenosis.

While nephrogenic hypertension is rather common in this condition, in few patients focal glomerulosclerosis and renal amyloidosis have also been described [11]. Lung involvement with necrotising pneumonia (lethal) has been reported in one patient [11].

The blood tests usually reveal an elevation of acute phase reactants (ERS, CRP), low haemoglobin levels and neutrophilic leukocytosis [1, 2]; however in few patients cytopenia (pancytopenia, leucopoenia) has been detected [1, 7, 12]. Auto-antibody are usually negative.

As stated above, a mild immunodeficiency can be observed; some patients present hypogammaglobulinemia that may affect IgM or all Ig subclasses [1, 13]. Of note, despite the low immunoglobulins’ levels, only few cases displayed an increased susceptibility to infections, that was rather severe in exceptional cases [1, 3, 7, 12, 13].

MRI is the most useful tool in the diagnosis of cerebral strokes; in fact CT scan as well as conventional angiography may not detect the smaller lacunar strokes and therefore underestimate the entity of involvement of the CNS [1].

Some patients underwent an angiographic investigation, that revealed the presence of stenosis and/or aneurysms of abdominal artery, particularly mesenteric, celiac, hepatic and renal arteries; the histological analysis, when done, revealed a necrotizing vasculitis [1, 2].

In patients with symptoms suggestive for organ involvement but without pathologic finding in not-invasive radiologic studies, conventional angiography can be of help revealing aneurism and or stenosis in the middle sizes arteries.

Skin biopsy revealed, in most cases, a non-granulomatous, necrotizing vasculitis of small and medium-sized vessels, with the same histopathologic features of polyarteritis nodosa [1, 2, 9].

In few cases the histology was less specific showing a leucocytoclastic vasculitis or a panniculitis.

Polyarteritis nodosa (PAN) is, according to the Chapel Hill classification, a “Necrotizing arteritis of medium or small arteries without glomerulonephritis or vasculitis in arterioles, capillaries, or venules, and not associated with antineutrophil cytoplasmic antibodies (ANCAs)” [22]. It’s gathered in the medium-sized vessels vasculitis, even if it can affect arteries of any size [22].

Being DADA2 a vasculitis with a genetic basis, it has been proposed to group this disease in the vasculitis with a probable cause according to the Chapel Hill classification [11, 22].

Notably, most of the DADA2 patients not only received the histological diagnosis of PAN but also met the EULAR/PRINTO/PRES diagnostic criteria for childhood polyarteritis nodosa (Table 4) [23].

Even if most of the patients with DADA2 have a clinical phenotype consistent with a systemic inflammatory vasculopathy, a recent report has enlighten that the disease may be dominated by clinical manifestations suggestive for an immune-disrective condition, such as cytopenia, lymphadenopathy, hepatosplenomegaly and immunodeficiency with severe viral infections [7]. The two patients described did not present skin involvement and one of them developed a vascular involvement only after bone-marrow transplantation. Of note, the mutations found in these two patients were the same described in patients with a “typical” inflammatory clinical picture.

Similarly a third patient with a lymphoprolipherative clinical picture, resembling Castleman’s syndrome, has been reported by the same group [5].

A more recent clinical series of 9 DADA2 patients with the homozygous R169Q mutation has enlightened that the presence of cytopenia is a common finding of the disease, together with the common inflammatory manifestations [12].

In the two patients carrying homozygous 22q11.1 deletion, encompassing both copies of the IL-17 receptor A (IL17RA) and the CECR1 gene, the clinical phenotype was dominated by muco-cutaneous infections and dermatitis associated to persistent inflammation and, in one patient, vasculitis responding to steroids [14]. Livedo reticularis, stroke and other DADA2 clinical manifestations were not described.

Finally two brothers with a clinical picture consistent with the diagnosis of common variable immunodeficiency (CVID) were found to be affected by DADA2 by whole exome-sequencing; of note only one of them displayed clinical sign and symptoms consistent with a vasculopathy [13].

Being a disease of recent identification, the clinical outcome has not been well investigated. However, from the clinical data by now available is clear that the spectrum of severity of the disease is wide, ranging from patients with neonatal onset and a severe organ involvement to patients with onset in the adulthood and the presence of only skin manifestations (Tables 2 and 3); of note, even between patients carrying the same mutations in CECR1 gene the clinical picture can be widely different (Tables 2 and 3).

The disease turned out to be lethal in seven out of the 65 patients by now described [1, 2, 6, 13, 14]: in three cases the severity of the visceral involvement was lethal [1, 2], two patients died for respiratory complications following intracranial haemorrhage [6, 13], while one patient developed necrotising pneumonia [1, 11]; finally one of the two patients carrying the homozygous deletion on 22.11.1 chromosome died for septic shock.

DADA2 is a newly recognised condition and the number of patients so far described is limited; for this reason the response to treatment is largely anecdotal and still controversial (Table 5).

Being an inflammatory condition, high doses of steroids are usually able to control the clinical manifestations [1, 2, 8, 9, 11, 12]. However, due to the severity of the condition, a steroid-dependence is often described. None of the most common immunosuppressive drugs (cyclophosphamide, azathioprine, methotrexate) was effective [1, 2, 6, 8, 11, 13].

Navon et al. reported ten patients treated with anti-TNF drugs (etanercept, adalimumab, infliximab) with complete response in 8, even after the failure of immunosuppressive therapies [2]; good results with anti-TNF agents were also reported in other small series [3, 11, 12]. By now, the reason why this drug is effective is still unclear.

According to the report of Zhou et al., neither immunosuppressive nor biologic drugs were able to completely control the disease manifestations in all treated patients; the enzymatic substitutive treatment (fresh frozen plasma or recombinant enzyme) was postulated to be of help. This approach was tempted in two patients reported by Batu et al. with a transient good response in one and a not-satisfactory response in the other [11].

A possible role of hematopoietic stem cell transplantation (HSCT) has been postulated to be effective by Zhou et al. and Navon et al., being able to provide ADA2 producing monocytes and therefore to normalize the plasmatic levels of the enzyme [1, 2]. This therapeutic strategy, performed in one of the two patients reported by Van Eyck et al. [7] and in a patient reported by the NIH group [3], was able to normalize the plasmatic levels of ADA2 and to control the disease manifestations [3, 7]; early complications occurred in one of them. More recently two additive patients who displayed a complete response to HSCT have been described [12].

Van Eyck et al. conclude that HSCT should be suggested only for those patients with a severe disease, since DADA2 patients present an increased risk of HSCT-related complications due to the persistent inflammation and the compromised endothelial integrity [7]. Of note, the other patient described in this paper displayed a complete response to treatment with sirolimus; the authors assume that this drug may be of help in the control of the clinical manifestations related to ADA2-deficiency, being able to reduce the M1 macrophage differentiation and the production of IL-6 [7].