Biomarkers in Hereditary Angioedema

Functional C1-INH (fC1-INH, C1-INH activity) plasma levels are considered a reference test in the diagnosis of HAE due to C1-INH deficiency (C1-INH-HAE), particularly of type 2 [14, 18‐20]. However, there have been some concerns in relation to this test, including variability due to sample handling and time of storage or an increase in its levels in the course of inflammation and infection [21‐23]. In addition, different types of assays may yield divergent results [19, 24]. Moreover, fC1-INH measurements can be affected by the presence of autoantibodies to C1-INH in the sample [16], as well as by danazol treatment [25] or plasma-derived C1-INH [26]. Currently, fC1-INH levels are measured by a commercial chromogenic assay or an enzyme-linked immunosorbent assay (ELISA), which detects complexes formed by C1-INH and complement C1s, activated FXII, or plasma kallikrein (PKa) [18, 21, 27]. However, functional assays have limited ability in detecting small changes over time [18, 21]. Recently, a highly robust point-of-care test to measure fC1-INH in dried blood spot has been described [28]. It is based on the quantitation of the enzyme reaction product by liquid chromatography–tandem mass spectrometry and is characterized by high reproducibility and accuracy, together with sample storage stability. Whereas the diagnostic value of functional C1-INH levels is well established, its value as a monitoring biomarker remains unclear. A predominant opinion is that the level of fC1-INH has little relationship to the clinical course of C1-INH-HAE [13, 14]. Suffritti et al. showed lower fC1-INH in patients during attacks than during remission (31 and 131 patients, respectively) [29]. However, Kajdasci et al. and Cugno et al. did not confirm this finding in their studies involving 18 and 28 patients with C1-INH-HAE, respectively [30, 31]. Other studies focused on associations between C1-INH function and disease course. Kelemen et al. demonstrated that the baseline level of fC1-INH correlates with the HAE severity score [32]. In a subsequent report by the same group, lower fC1-INH levels were observed in patients with a higher number of attacks and a higher need for C1-INH on-demand treatment [33]. In turn, Bafunno et al. could not find a significant correlation between fC1-INH level and disease severity score or age of onset [34]. Further insight was provided by clinical trials with plasma-derived C1-INH supplementation. A level of approximately 40% of fC1-INH appears to protect against angioedema attacks in most patients who received prophylactic treatment with subcutaneous C1-INH [35], which is in line with previous clinical observations [36, 37]. Therefore, these ranges of fC1-INH may serve as a prognostic biomarker of disease activity, i.e., assessment of likelihood of future attacks.

Antigenic C1-INH (AgC1-INH) plasma concentration is a critical diagnostic biomarker for the diagnosis of C1-INH-HAE type 1 [38]. Its concentration can be measured by nephelometry, turbidimetry, or radial immunodiffusion, depending on local availability and cost [23]. The results may be influenced by replacement therapy with C1-INH [14]. In principle, AgC1-INH is not considered a valuable monitoring biomarker of the clinical course of the disease [13, 14, 30, 31]. However, Spath et al. demonstrated most frequent attacks in patients with C1-INH-HAE when AgC1-INH levels were below 0.035 g/l [39]. Other authors reported that AgC1-INH levels were lower during attacks [29] or negatively correlated with the annual number of attacks [33].

The protease-inhibitor complex C1-INH-C1(r,s) reflects contact system activation and thus may be considered a potential biomarker [14]. Its plasma concentration, which can be determined by ELISA, is influenced by the amount of C1-INH present in plasma and is artificially low in C1-INH deficiency [14, 33]. Plasma levels of C1-INH-C1(r,s) complexes were found to be higher in patients with C1-INH-HAE than in healthy controls [33, 40] and increased further during angioedema attacks [41]. Patients with higher C1-INH-C1(r,s) levels had a history of more severe attacks and more often required emergency treatment [40]. Plasma C1-INH-C1(r,s) complex levels were reported to normalize in patients treated with stanozolol along with a reduction in symptoms [42], thus showing promise as a biomarker for monitoring therapeutic response (a response biomarker) [43].

Complement C4 serves as an important contributory diagnostic biomarker in C1-INH-HAE [14, 38], because its level is reduced in most patients, especially during attacks [16, 24]. Nevertheless, its performance as a monitoring biomarker is poor. Complement C4 was shown to correlate with the frequency of attacks and on-demand consumption of C1-INH concentrate [33] but not with disease severity scores [32]. No significant difference was observed in C4 levels between remission and acute abdominal attacks [31].

Additional laboratory complement indices were also reported in HAE. Varga et al. found that anti-C1-INH IgM antibody levels correlated with disease severity in C1-INH concentrate–naive patients [44]. Other reports indicated that the levels of Mannose-binding lectin-associated serine proteases (MASP-1, MASP-1)-C1-INH complexes are lower in C1-INH-HAE patients and correlate with the frequency of attacks [45], whereas the levels of MASP-2 and ficolin-3/MASP-2 complexes increase during attacks [46] (Table 1).

Since bradykinin (BK) is the major mediator of swellings in HAE [13, 47], it is expected to be the most accurate biomarker of upcoming attacks. Indeed, plasma BK levels were reported to be higher in patients with C1-INH-HAE than in healthy controls, with a further significant rise during attacks [48]. Moreover, they were found higher in blood taken from the site of angioedema than at a control site in classical report by Nussberger et al. [49]. Nevertheless, clinical utility of plasma BK levels is being questioned because of its high sensitivity to pre-analytical procedures and a very short half-life (measured in seconds) [13, 14]. In addition, the measurement of BK and its breakdown metabolites, such as des-Arg-BK, based on liquid chromatography with mass spectrometry, is technically challenging [50, 51]. The clinical utility of a commercially available assay kit for detecting the product of BK degradation, which might be a surrogate for BK quantification, awaits confirmation [13].

High molecular weight kininogen (HK) proteolysis by active PKa results in generation of cleaved HK (cHK) and BK. Thus, cHK is thought to be a promising indirect marker of BK release and contact system activation, which occurs during HAE attacks. Western blotting for cHK detection is hard to standardize and subjective in interpretation. Semi-quantification of results can be done by scanning gels [13]. New effective methods for cHK measurement are based on a monoclonal antibody ELISA [52] and modification of liquid chromatography–mass spectrometry [53]. Cugno et al. observed, in patients with C1-INH-HAE, high levels of cHK during attacks and normal levels during remission [54, 55]. In a subsequent report on a large population of C1-INH-HAE patients, Suffritti et al. reported cHK levels to be higher in patients than in controls and to further increase during attacks. They also demonstrated that cHK levels were higher in highly symptomatic patients in comparison with those with less frequent attacks [29]. Using an immunoassay based on ELISA, Hofman et al. confirmed that cHK levels were elevated in patients with C1-INH-HAE during remission and further increased during attack [52]. Banerji et al. showed significant reductions in cHK levels in C1-INH-HAE patients treated with 300 mg and 400 mg of lanadelumab, a monoclonal antibody against PKa [56]. In turn, Bova et al. measured cHK levels in a cohort of 105 HAE patients with normal C1-INH levels (nC1-INH-HAE), including U-HAE and FXII-HAE [57]. cHK was measured during remission in plasma collected with and without using protease inhibitors. In patients with U-HAE, cHK levels were similar to those in healthy controls with the use of protease inhibitors and significantly increased without them. In patients with FXII-HAE without the use of protease inhibitors, cHK levels were higher than in controls and similar to those observed in patients with U-HAE [57].

Therefore, cHK levels can discriminate between patients and healthy individuals, between highly symptomatic and less symptomatic patients, and between acute attack and remission. The development of a new reliable and less laborious measurement method may facilitate the clinical application of this parameter.

Plasma kallikrein (PKa) is also considered a potential biomarker of BK-mediated angioedema attacks. Spontaneous PKa activity, measured with the use of a chromogenic substrate, was higher in patients with C1-INH-HAE than in controls and further increased during attacks [29]. Similar results were reported by other authors [58]. Lara-Marquez et al. developed an assay based on plasma capacity to generate active PKa under ex vivo stimulation with dextran sulfate [59]. Using different threshold cutoff values in this assay, patients with BK-dependent angioedema (nC1-INH-HAE and C1-INH-HAE) could be distinguished from controls without swelling attacks and patients with histaminergic angioedema, based on a higher amount of generated PKa [59].

Activated coagulation factor FXII (FXIIa) was shown to be higher in patients with C1-INH-HAE than in healthy controls, with a further increase during attacks [54, 58, 60]. The activity of FXII was also evaluated as a possible biomarker of FXII-HAE in symptom-free periods, but studies provided inconsistent results, with some suggesting an increase while others reporting no difference between patients and healthy controls [61, 62].

Factor XIIa/C1-INH complexes were studied by Konings et al. [63]. They showed that the levels of these complexes, as well as the levels of FXIa-C1INH and PKa-C1INH complexes, were lower in patients with C1-INH-HAE than in healthy controls after an in vitro activation of the samples with an FXII trigger.

Kinins—degradation and accumulation of kinins, including BK, may modify the clinical phenotype of HAE. Carboxypeptidase N (CPN), angiotensin-converting enzyme (ACE), and aminopeptidase P (APP) are major kinases involved in kinin catabolism [64]. Drouet et al. investigated the kininase activity in C1-INH-HAE patients with and without androgen prophylaxis [65]. APP levels were inversely correlated with disease severity in all patients, and CPN activity showed the same relationship only in untreated patients. Subsequently, a similar study was conducted in patients with FXII-HAE, showing an inverse correlation between the disease severity score and both ACE and CPN activities, but not APP [66]. Moreover, the total activity of serine proteases was shown to be higher in patients with C1-INH-HAE and nC1-INH-HAE than in healthy controls and to further increase in those with nC1-INH-HAE during attacks [67].

Coagulation and fibrinolysis are also pathways activated in HAE and therefore extensively investigated for potential biomarkers. To date, a wide range of fibrinogen split products in patients during HAE attack and remission have been studied [31, 55, 60, 68, 69]. The findings from these studies are summarized in Table 1.

Plasminogen activator inhibitor (PAI)-1 levels, prothrombin time, and activated partial thromboplastin time were also found to be lower during C1-INH-HAE attacks compared with a symptom-free period [60]. The PAI-1 level was also found lower in patients with U-HAE and FXII-HAE compared with control individuals, but the difference was not significant [70]. In the same study, also PAI-2 levels were reported to be lower in patients with U-HAE and FXII-HAE than in controls, but this finding was not corroborated by other studies [71].

D-dimer levels were also found elevated and may also discriminate between abdominal HAE attacks and abdominal colic episodes [31, 72], between multiple- and single-site attacks [60], and between submucosal (abdominal, oropharyngeal–laryngeal) and subcutaneous (peripheral, facial) attacks [73]. D-dimer levels were observed to decrease at day 7 after attack [72, 73]. Prophylaxis with continuous supplementation of subcutaneous plasma-derived C1-INH seems to suppress D-dimer levels, in parallel with a significant reduction in the rate of HAE attacks [74].

Collectively, laboratory parameters associated with coagulation and fibrinolysis may have limited usefulness as biomarkers because of their interpatient variation in the plasma [13]. They can be useful to some extent for disease course monitoring when compared with baseline values of the same individual. Another major disadvantage that limits their potential use as biomarkers is the fact that they may not be specific for HAE. For instance, a D-dimer test is widely used and easily available in emergency settings, but D-dimer levels are elevated in numerous medical conditions, i.e., pulmonary embolism and venous thromboembolism, but also in infection, autoimmune disorders, malignancy, as well as in smokers, pregnant women, and elderly individuals [75].

As the endothelium is inherently related to microvascular permeability and, in consequence, the swelling phenomenon, it represents another area of interest regarding potential biomarkers in HAE. Studies in this field involved vascular endothelial cadherin (transmembrane adhesive protein) [72, 76], von Willebrand factor (marker of endothelial damage), soluble E-selectin (cytokine-induced adhesion molecule), endothelin-1 (vasomotor activity regulator) [30, 77, 78], arginine vasopressin, adrenomedullin [78], atrial natriuretic peptide [79], as well as endothelial-derived endocan and vascular cell adhesion molecule-1 (markers of endothelial function) [80]. The subsequently studied modulators of vascular permeability included vascular endothelial growth factors, angiopoietin-1 (which promotes endothelial stabilization) and angiopoietin-2 (which facilitates vascular permeability) [81, 82], secreted phospholipases A2 (particularly the 2A group) [83], and platelet-activating factor acetylhydrolase [81]. Studies investigated changes in the above factors in small groups of patients with C1-INH-HAE during attacks and symptom-free periods, in comparison with healthy controls, as shown in Table 1. More recent research also involved patients with nC1-INH-HAE in remission. Bova et al. reported increased levels of angiopoietin-1 and vascular endothelial growth factors A and C in patients with U-HAE as well as increased levels of vascular endothelial growth factor C in patients with FXII-HAE [57].

It is well known that stimulation of endothelial cells with BK and its analogs causes the release of prostacyclin and unstable endothelium-derived relaxing factor [84], which was subsequently identified as nitric oxide species contributing to enhanced vascular permeability [85]. Bas et al. investigated the long-lived metabolite of prostacyclin, namely, 6-keto-prostaglandin F1-α, as a potential biomarker for the diagnosis of angioedema induced by angiotensin-converting enzyme inhibitors [86]. In turn, Demirturk et al. showed that plasma levels of endothelial nitric oxide synthase were significantly higher in patients with C1-INH-HAE in remission and during attacks than in healthy participants and that the levels of nitric oxide metabolites were elevated only during attacks [87]. Further research involved also the effect of HAE on endothelial function expressed by blood concentrations of asymmetric dimethylarginine, a strong inhibitor of nitric oxide synthesis associated with numerous common conditions, such as atherosclerosis. In a group of HAE patients (24 with C1-INH-HAE and 14 with FXII-HAE), asymmetric dimethylarginine levels were higher than in controls [88]. In a subsequent report, this group demonstrated that serum concentrations of advanced oxidation protein products, used as markers of oxidative stress, were higher in patients with C1-INH-HAE and FXII-HAE during remission than in controls [89]. In another study investigating oxidative stress in HAE, the authors showed that the levels of reactive oxygen species in peripheral blood mononuclear cells of patients with C1-INH-HAE were higher than in controls; however, no differences were observed between the groups in plasma levels of advanced oxidation protein products [90].

A number of studies have investigated the association of HAE with low-grade inflammation markers, immune system elements, and hormones. Investigators compared patients with C1-INH-HAE and healthy controls with respect to a number of factors, including C-reactive protein, erythrocyte sedimentation rate, white blood cell, and neutrophil counts [91‐93], a wide network of pro-inflammatory and anti-inflammatory cytokines (i.e., interleukins IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-13, IL-17, interferon-γ, tumor necrosis factor-α, granulocyte colony-stimulating factor, and granulocyte-macrophage colony-stimulating factor) [87, 94, 95], as well as sex hormones, such as progesterone, together with sex hormone-binding globulin [96]. Differences in these factors between patients during remission and attacks, as well as in comparison with healthy controls, are summarized in Table 1. More recently, significant changes were shown in multifunctional human glycoprotein fetuin-A levels between symptom-free periods and attacks in patients with C1-INH-HAE [93]. Moreover, fragmentation patterns of serum glycoprotein 120 were also reported as a potentially useful biomarker in patients with nC1-INH-HAE [97] (Table 1).

The utility of biomarkers reported in the above studies should be interpreted with caution if (i) biochemical parameters investigated in the study are involved also in other clinical pathologies (low specificity in HAE, uncertain direction of a causative relationship between the studied laboratory parameter, and HAE); (ii) the study includes a small sample size of patients; and (iii) results of laboratory measurements are highly variable. For biomarkers expected to facilitate the monitoring of HAE severity or response to treatment, well-defined endpoints are necessary. However, differences in symptom severity scores used and lack of standardization make it difficult to compare the results between studies. In turn, potential diagnostic biomarkers for HAE or angioedema attacks require a gold standard for their validation. Hence, a new test based on the measurement of a given molecule has to be validated against proven clinical events or cases (e.g., an investigator-confirmed attack or a known genetic mutation with proven pathogenicity). Studies designed to elucidate the mechanisms of HAE usually are unable to strictly follow the above requirements for biomarker identification, but there are already some ongoing trials directly dedicated to the study of biomarkers in HAE, including specific biomarkers of BK-mediated angioedema attack in a pediatric population, potential disease-specific biomarkers for HAE with the use of dry blood spot technology, and, finally, the use of cHK as a biomarker in C1-INH-HAE [1].